Psychology
holiday Jessica
Larzelere-Kuhn2005_Article_ComparingChildOutcomesOfPhysic.pdf
Clinical Child and Family Psychology Review, Vol. 8, No. 1, March 2005 ( C© 2005) DOI: 10.1007/s10567-005-2340-z
Comparing Child Outcomes of Physical Punishment and Alternative Disciplinary Tactics: A Meta-Analysis
Robert E. Larzelere1,2 and Brett R. Kuhn1
This meta-analysis investigates differences between the effect sizes of physical punishment and alternative disciplinary tactics for child outcomes in 26 qualifying studies. Analyzing differences in effect sizes reduces systematic biases and emphasizes direct comparisons be- tween the disciplinary tactics that parents have to select among. The results indicated that effect sizes significantly favored conditional spanking over 10 of 13 alternative disciplinary tactics for reducing child noncompliance or antisocial behavior. Customary physical punish- ment yielded effect sizes equal to alternative tactics, except for one large study favoring phys- ical punishment. Only overly severe or predominant use of physical punishment compared unfavorably with alternative disciplinary tactics. The discussion highlights the need for better discriminations between effective and counterproductive use of disciplinary punishment in general.
KEY WORDS: children; parenting; discipline; punishment; spanking.
Uncertainty about the effects of physical pun- ishment on children has persisted despite decades of research. Two major perspectives have emerged recently. The first is an unconditional anti-spanking perspective, advanced by both social scientists (Gershoff, 2002; Straus, 2001) and advocacy groups (EPOCH-Worldwide, 2004). In response, at least 13 countries have passed laws banning all physical pun- ishment by parents (EPOCH-Worldwide, 2004).
The second perspective, which has been called the conditional-spanking perspective (Benjet & Kazdin, 2003), has attempted to identify condi- tions under which spanking may be beneficial or at least not detrimental to children. The conditional- spanking perspective emphasizes the parenting con- text and manner of implementation, which may distinguish effective from counterproductive uses of punishment more than its form (e.g., physical or nonphysical). In one sense the disciplinary ac-
1Psychology Department, Munroe-Meyer Institute, University of Nebraska Medical Center, Nebraska.
2Address all correspondence to Robert E. Larzelere, Psychology Department, MMI, 985450 Nebraska Medical Center, Omaha, Nebraska 68198-5450; e-mail: [email protected].
tions of parents in most cultures have reflected a conditional-spanking perspective until recently. In 1994–1995, for example, 94% of American parents and 52% of Canadian parents of 3- and 4-year-olds reported using physical punishment at least occasion- ally (Larzelere, 2004; Straus & Stewart, 1999). The conditional-spanking perspective holds that spank- ing should be investigated under the conditions for which parents have considered it advisable before imposing a spanking ban on parents (Bauman & Friedman, 1998; Baumrind, Larzelere, & Cowan, 2002; Eysenck, 1993; Friedman & Schonberg, 1996b; Larzelere, Baumrind, & Polite, 1998).
Two recent literature reviews from these two perspectives did little to resolve the issue. Gershoff’s (2002) meta-analysis concluded that physical pun- ishment was linked positively to immediate com- pliance, but negatively with 10 other outcomes in children and families. In a qualitative review, Larzelere (2000) concluded that causal evidence showed that nonabusive spanking of 2–6-year-olds produced more beneficial than detrimental child out- comes when it was used to enforce milder disci- plinary tactics such as reasoning or time-out, es- pecially in subcultural groups that support its use.
1
1096-4037/05/0300-0001/0 C© 2005 Springer Science+Business Media, Inc.
2 Larzelere and Kuhn
Benjet and Kazdin (2003) recently compared the two reviews and concluded, “A top priority for re- search on spanking would seem to be a comparison of spanking with alternative procedures that already have considerable evidence in their behalf” (p. 215). The current meta-analysis attempts to address this priority by investigating the studies included in either review that examined one or more alternative dis- ciplinary tactics in addition to physical punishment. It also investigates several methodological problems that could explain the discrepant conclusions from the two reviews (Benjet & Kazdin, 2003).
To provide a context for this meta-analysis, we briefly summarize the methodological problems that have hindered definitive conclusions about physical punishment. We then clarify why a meta-analysis using differences in effect sizes between physical punishment and disciplinary alternatives can reduce these methodological problems.
Methodological Issues
The spanking controversy persists largely be- cause pervasive methodological problems have per- mitted a wide range of interpretations. These prob- lems include predominantly correlational research; failing to discriminate among nonabusive, customary, and overly severe use of physical punishment; mea- suring disciplinary practices and child outcomes from the same information source; and failing to rule out plausible alternative explanations.
The strongest evidence against physical punish- ment in Gershoff’s (2002) thorough meta-analysis consisted of longitudinal correlations, i.e., zero-order correlations between physical punishment and sub- sequent child outcomes. Although such correlations are consistent with a causal effect (Smith, 2002), their pattern is typical of most corrective interven- tions (Larzelere, Kuhn, & Johnson, 2004). In post- treatment comparisons, recipients of corrective inter- ventions will compare poorly to those not needing such interventions, whether the intervention is deliv- ered by physicians (e.g., radiation treatment), educa- tors (Head Start), psychologists (marital counseling), or parents (punishment).
Consider radiation treatment as an example. Pa- tients who received radiation treatment last year are more likely to experience cancer-related symptoms this year than those who did not receive (or need) radiation treatment. Longitudinal zero-order corre- lations would indicate that radiation treatment is
associated with increased cancer-related symptoms. Of course, the initial presenting problem (cancer) is the causal factor underlying that correlation be- cause it leads to both the corrective intervention (radiation treatment) and the subsequent outcome (cancer). Consequently, zero-order longitudinal cor- relations cannot discriminate effective corrective in- terventions from those that are counterproductive.
Second, most of the research on physical pun- ishment “lumps” together nonabusive and custom- ary punishment with overly severe forms of physi- cal punishment. For example, 65% of the studies in Gershoff’s (2002) meta-analysis included overly se- vere physical punishment in their measure, accord- ing to Baumrind et al. (2002). Examples ranged from vaguely defined “punitive discipline” (6% of the studies), composite measures of the frequency and severity of physical punishment (29%), and the in- clusion of extreme violence (31%), such as slapping in the face (seven studies), beating up (three studies), or hitting with a fist and causing bruises and cuts (one study).
Third, many studies of disciplinary tactics have based the antecedent and consequent variables on the same source of information. Typically, moth- ers reported both their disciplinary tactics and their child’s behavior. In retrospective studies, grown chil- dren reported both their current functioning and the disciplinary tactics they received earlier in life. This same-source bias has been shown to inflate associ- ations between disciplinary tactics and adverse out- comes (Yarrow, Campbell, & Burton, 1968).
Finally, plausible alternative explanations of the data on physical punishment have not been ruled out, resulting in widely discrepant explanations for the varied outcomes across studies. Consider the strongest evidence of the effectiveness of spanking. Four small-randomized clinical studies found that spanking was effective in reducing defiance in clin- ically oppositional 2–6-year-olds (Bean & Roberts, 1981; Roberts; Day & Roberts, 1983; Roberts, 1988; Roberts & Powers, 1990). The difference in effect sizes between those four randomized studies (mean d = 1.21) and the 113 non-randomized studies (mean d = −.35) in Gershoff (2002) approached the largest difference ever found in a meta-analysis (Lipsey & Wilson, 1993). This difference could be explained by one or more of the following confounded interpre- tations. Compared to the non-randomized studies, Roberts’ four randomized studies (1) had causally stronger evidence, (2) limited spanking to two open- handed swats under the supervision of a clinical
Physical Punishment vs. Alternative Tactics 3
psychologist, (3) used spanking only to enforce com- pliance with time-out, (4) applied only to children from 2 to 6 years of age who (5) were clinically re- ferred for oppositional behavior problems, and (6) focused on decreases in defiance in the clinic as the primary outcome. Whereas advocates of the anti- spanking viewpoint consider the type of outcome (short-term compliance) to be the crucial distinction (Gershoff, 2002; Straus, 2001), conditional-spanking researchers emphasize the stronger causal evidence, the specific conditions in the randomized studies (e.g., the child’s age, the discipline situation), and the way in which spanking was implemented (Baumrind et al., 2002; Larzelere, 2000).
Although these four methodological problems are often acknowledged, the extent to which they undermine research conclusions has received insuf- ficient attention. Suppose radiation treatment were studied in the same way that researchers have investi- gated physical punishment. Borrowing statistics from Gershoff’s (2002) thorough meta-analysis, two-thirds (65%) of studies of radiation treatment would have included excessive dosages of radiation (Baumrind et al., 2002), 58% would have been cross-sectional stud- ies, and only 4% would have taken into considera- tion the presence or severity of cancer. Would it be surprising that patients who received radiation treat- ment last year had higher rates of cancer both last year and this year, compared to those who did not receive (or need) radiation? A meta-analysis of ra- diation treatment using predominantly correlational studies would come to the same conclusions as Ger- shoff’s (2002) meta-analysis, specifically that radia- tion treatment is consistently linked to detrimental outcomes. As aptly noted by Straus (2001), valid causal conclusions require controlling for the effects of initial child misbehavior. Otherwise, initial child misbehavior may lead to more disciplinary tactics as well as worse child outcomes, which would account for the associations found by Gershoff (2002).
Rationale for a Meta-Analysis of Differential Effect Sizes
This meta-analysis attempts to reduce these per- vasive methodological problems by (1) distinguishing among four types of physical punishment, (2) basing effect sizes on each study’s strongest methodological evidence whenever possible, and (3) analyzing differ- ential effect sizes between physical punishment and alternative disciplinary tactics.
To address the “lumping” problem, we distin- guish among conditional spanking, customary phys- ical punishment, overly severe physical punishment, and predominant use of physical punishment. Con- ditional spanking (as labeled by Benjet & Kazdin, 2003) refers to spanking under the limited condi- tions that have been associated with better child out- comes (e.g., spanking when a 2–6-year-old refuses to comply with time out). The purpose of distinguish- ing this category is to determine whether spanking is associated with better outcomes than alternative tactics even under ideal conditions. Customary phys- ical punishment represents the manner in which par- ents typically use physical punishment. The purpose of this category is to investigate whether typical use of physical punishment is associated with better or worse outcomes than alternative tactics. Overly se- vere physical punishment includes the use of exces- sive force, hitting with an object, or slapping in the face (Baumrind et al., 2002). Finally, predominant us- age indicates that physical punishment is the parent’s primary disciplinary method, i.e., it is preferred over milder disciplinary tactics.
This meta-analysis bases effect sizes on the findings from each study that are methodologically strongest. For example, our effect sizes are based on results that take initial child misbehavior into ac- count from distinct sources of information, when- ever possible. This choice contrasts with Gershoff’s (2002) decision to base effect sizes on correlations for the sake of consistency, ignoring methodologi- cally stronger findings in several studies.
Finally, this meta-analysis estimates differences in the effect sizes of physical punishment vs. alter- native disciplinary tactics, using identical methods within the same study. If the apparently detrimen- tal child outcomes reflect causal effects unique to physical punishment, then the effect sizes of physi- cal punishment should compare poorly to the effect sizes of alternative disciplinary tactics. On the other hand, if detrimental child correlates of physical pun- ishment represent methodological artifacts, then the effect sizes of alternative disciplinary tactics should appear equally detrimental.
A methodology for analyzing differences be- tween effect sizes is already well established for ran- domized studies. It is based on the differential ef- fect size contrasting post-treatment outcomes from a treatment and a control group. For the usual effect size measure (d), this is the same as calculating an effect size for each group (e.g., improvement from pre-to-post) and then using the difference between
4 Larzelere and Kuhn
those two effect sizes. This equality is based on two assumptions. First, the treatment and control group must have identical pre-treatment scores, which ran- domization guarantees in the long run.3 The second assumption is that the effect sizes for the treatment and control groups are based on the same stan- dard deviation. When these assumptions apply, typi- cal meta-analyses of randomized clinical trials can be considered equivalent to analyses of differences be- tween effect sizes. The only distinction in our meta- analysis is that it compares two treatments (disci- plinary tactics) with each other rather than treatment and control groups.
The major advantage of analyzing differences between effect sizes, however, is for non-randomized studies, which dominate this literature. Causal con- clusions can be supported from correlational stud- ies only to the extent that plausible alternative in- terpretations have been ruled out (Larzelere et al., 2004; Shadish, Cook, & Campbell, 2002). This prin- ciple applies to individual studies as well as to meta- analyses. For example, an alternative explanation for the positive correlation between physical punishment and subsequent antisocial behavior is that the child’s initial antisocial behavior may increase both the fre- quency of physical punishment and subsequent an- tisocial behavior. Just as individual studies control for this possibility by using initial child misbehavior as a covariate, this meta-analysis uses differences be- tween effect sizes to control for initial child misbe- havior.
A second advantage of analyzing differences be- tween effect sizes is that they allow researchers to directly compare realistic disciplinary choices. Bas- ing effect sizes on simple associations between a dis- ciplinary tactic and a child outcome implicitly com- pares parents who use that disciplinary tactic with those who do not use it. Instead of choosing between a given disciplinary tactic and doing nothing, parents typically choose between two or more alternative dis- ciplinary responses (Ritchie, 1999). Differences in ef- fect sizes are better suited for such comparisons.
In summary, this meta-analysis uses differences between effect sizes to control for confounds that influence all disciplinary tactics, e.g., selection bias due to initial child misbehavior. It falls short of be- ing causally definitive, however, because it rules out only some plausible interpretations of the underly- ing empirical evidence. This strategy is a substantial
3When pre-test scores differ, relative pre-post gains provide a fairer comparison than post-treatment differences.
improvement over typical meta-analytic methods for correlational data because it controls for important confounds and rules out alternative interpretations associated with them. At the very least, the cur- rent meta-analysis can determine whether the corre- lationally based effect sizes are uniquely detrimen- tal for physical punishment, are more detrimental for some disciplinary tactics than others, or are equally detrimental for all disciplinary tactics. Making these distinctions is a crucial step toward designing more causally informative studies in the future. The re- sults also have important implications for how phys- ical punishment should be used, if at all, and which alternative disciplinary tactics might be used instead.
METHOD
Literature Selection
Research studies were selected for this meta- analysis from recent reviews by Gershoff (2002) and Larzelere (2000). Both reviews attempted to be ex- haustive within their inclusion criteria for at least the previous 26 years. Additional selection criteria include the following: (1) The study must have inves- tigated one or more recommended alternative dis- ciplinary tactics as well as physical punishment, us- ing similar research methods. (2) The children had to average less than 13 years old at the time of the discipline. Most retrospective studies were excluded because they pertained to physical punishment of teenagers, based on the finding that retrospective re- ports of physical punishment correlated most highly with mothers’ reported physical punishment at 12– 14 years old (Stattin, Janson, Klackenberg-Larsson, & Magnusson, 1995). To be included, retrospective surveys had to ask specifically for disciplinary tactics at a younger age. (3) Selected studies had to inves- tigate at least one child outcome, excluding studies that investigated only parental outcomes.
These criteria yielded 26 studies that investi- gated physical punishment and one or more al- ternative tactics, summarized in Table I. Only eight of these studies were included in both previ- ous reviews. Eleven studies from Gershoff’s (2002) meta-analysis were excluded from Larzelere’s (2000) review because they were cross-sectional (seven studies) or used overly broad measures of punish- ment (three studies). One other study was incorrectly excluded from Larzelere’s (2000) review, because it did specify a younger age in its retrospective survey
Physical Punishment vs. Alternative Tactics 5 T
ab le
I. E
ff ec
t S
iz es
o f
P h
ys ic
al P
u n
is h
m en
t an
d A
lt er
n at
iv e
D is
ci p
li n
ar y
R es
p o
n se
s o
n C
h il
d O
u tc
o m
es
B as
is o
f ef
fe ct
si ze
& A
ge ,G
en d
er D
is ci
p li
n e
E ff
ec t
d is
cr ep
an ci
es fr
o m
S tu
d y
(P ar
en t)
a S
am p
le N
ta ct
ic b
O u
tc o
m e
si ze
c (d
) G
er sh
o ff
(2 00
2)
S h
o rt
-t er
m co
m p
li an
ce R
an d
o m
iz ed
C li
n ic
al T
ri al
s B
ea n
an d
R o
b er
ts (1
98 1)
2– 6
B ,G
(M )
C li
n ic
-r ef
er re
d fo
r d
is ru
p ti
ve b
eh av
io r
p ro
b le
m s
8 2-
sw at
sp an
k b
ac k
-u p
fo r
ti m
e- o
u t
(T O
; co
n d
it io
n al
p h
ys ic
al p
u n
is h
m en
t [P
P ])
O b
se rv
ed co
m p
li an
ce ra
te to
30 p
ar en
ta l
co m
m an
d s
2. 80
d G
ai n
fr o
m p
re -t
es t
to p
o st
-t es
t m
ea n
s. G
er sh
o ff
co m
p ar
ed p
o st
-t es
t m
ea n
s fo
r th
e co
m p
le te
p ar
en t
tr ai
n in
g p
ro ce
d u
re vs
.a n
o -t
re at
m en
t co
n tr
o l
gr o
u p
8 C
h il
d -d
et er
m in
ed re
le as
e fr
o m
ti m
e- o
u t
(T O
)
C o
m p
li an
ce 1.
07 G
ai n
fr o
m p
re -
to p
o st
-t es
t
8 2-
sw at
sp an
k b
ac k
-u p
fo r
ti m
e- o
u t
(T O
; co
n d
it io
n al
P P
)
T im
e- o
u t
su cc
es s
(T im
e- o
u ts
re q
u ir
ed d
u ri
n g
30 co
m m
an d
s)
1. 61
N u
m b
er o
f ti
m e-
o u
ts n
ee d
ed co
m p
ar ed
to th
e ch
il d
-d et
er m
in ed
re le
as e
fr o
m ti
m e-
o u
t. O
u tc
o m
e n
o t
in G
er sh
o ff
D ay
an d
R o
b er
ts (1
98 3)
2– 5
B ,G
(M )
C li
n ic
-r ef
er re
d fo
r d
is ru
p ti
ve b
eh av
io r
p ro
b le
m s
8 2-
sw at
sp an
k b
ac k
-u p
fo r
ti m
e- o
u t
(T O
; co
n d
it io
n al
P P
)
C o
m p
li an
ce ra
te to
30 p
ar en
ta lc
o m
m an
d s
3. 45
d G
ai n
fr o
m p
re -t
es t
to p
o st
-t es
t. G
er sh
o ff
co m
p ar
ed th
e p
o st
-t es
ts fo
r th
e sp
an k
vs .t
h e
b ar
ri er
b ac
k -u
p s
8 B
ar ri
er (r
o o
m T
O )
b ac
k -u
p fo
r ti
m e-
o u
t (T
O )
C o
m p
li an
ce 3.
51 G
ai n
fr o
m p
re -
to p
o st
-t es
t
R o
b er
ts (1
98 8)
2- 6
B ,G
(M )
C li
n ic
-r ef
er re
d fo
r b
eh av
io r
p ro
b le
m s
9 2-
sw at
sp an
k b
ac k
-u p
fo r
T O
(c o
n d
it io
n al
P P
)
C o
m p
li an
ce ra
te to
30 +
p ar
en ta
l co
m m
an d
s
1. 60
d G
ai n
fr o
m p
re -t
es t
to p
o st
-t es
t. S
tu d
y n
o t
in G
er sh
o ff
9 B
ar ri
er (r
o o
m T
O )
b ac
k -u
p fo
r T
O C
o m
p li
an ce
1. 45
G ai
n fr
o m
p re
- to
p o
st -t
es t
9 2-
sw at
sp an
k b
ac k
-u p
fo r
T O
(c o
n d
it io
n al
P P
)
N u
m b
er o
f ti
m e-
o u
ts (T
O s)
an d
T O
es ca
p es
p ri
o r
to co
m p
li an
ce to
10 co
m m
an d
s
.1 3
M ea
n n
u m
b er
o f
T O
s an
d T
O es
ca p
es co
m p
ar ed
to th
e b
ar ri
er b
ac k
-u p
R o
b er
ts an
d P
o w
er s
(1 99
0) 2–
6 B
,G (M
) C
li n
ic -r
ef er
re d
fo r
b eh
av io
r p
ro b
le m
s 9
2- sw
at sp
an k
b ac
k -u
p fo
r T
O (c
o n
d it
io n
al P
P )
C o
m p
li an
ce ra
te to
u p
to 60
p ar
en ta
l co
m m
an d
s
1. 34
c G
ai n
fr o
m p
re -t
es t
to p
o st
-t es
t. G
er sh
o ff
co m
p ar
ed th
e sp
an k
p o
st -t
es t
vs .t
h e
o th
er 3
b ac
k -u
p s
co m
b in
ed 9
B ar
ri er
(r o
o m
T O
) b
ac k
-u p
fo r
T O
C o
m p
li an
ce 2.
17 G
ai n
fr o
m p
re -
to p
o st
-t es
t
9 R
es tr
ai n
t b
ac k
-u p
fo r
T O
C o
m p
li an
ce 1.
15 S
am e
9 C
h il
d -d
et er
m in
ed T
O re
le as
e C
o m
p li
an ce
1. 52
S am
e
9 2-
sw at
sp an
k b
ac k
-u p
fo r
T O
(c o
n d
it io
n al
P P
)
T O
su cc
es s
(c h
il d
re n
w it
h o
u t
ex ce
ss iv
e T
O s
o r
ex ce
ss iv
e T
O es
ca p
es )
.4 4
P ro
p o
rt io
n o
f ch
il d
re n
w it
h su
cc es
sf u
l T
O co
m p
li an
ce ,c
o m
p ar
ed to
ch il
d -d
et er
m in
ed T
O re
le as
e gr
o u
p .
O u
tc o
m e
n o
t u
se d
b y
G er
sh o
ff 9
B ar
ri er
b ac
k -u
p fo
r T
O T
O su
cc es
s 1.
02 S
am e
6 Larzelere and Kuhn
T ab
le I.
C o
n ti
n u
ed
B as
is o
f ef
fe ct
si ze
& A
ge ,G
en d
er D
is ci
p li
n e
E ff
ec t
d is
cr ep
an ci
es fr
o m
S tu
d y
(P ar
en t)
a S
am p
le N
ta ct
ic b
O u
tc o
m e
si ze
c (d
) G
er sh
o ff
(2 00
2)
9 R
es tr
ai n
t b
ac k
-u p
fo r
T O
T O
su cc
es s
−. 22
S am
e
W it
h in
-s u
b je
ct se
q u
en ti
al an
al ys
es L
ar ze
le re
et al
.( 19
96 )
2– 3
B ,G
(M )
V o
lu n
te er
s 38
S la
p h
an d
o r
sp an
k ,
w h
et h
er re
as o
n in
g w
as al
so u
se d
o r
n o
t (c
u st
o m
ar y
P P
, co
n tr
as te
d w
it h
n ex
t tw
o ta
ct ic
s)
D el
ay u
n ti
l n
ex t
re cu
rr en
ce o
f d
is o
b ed
ie n
ce ,
co m
p ar
ed to
ty p
ic al
d el
ay s
fo r
th at
ch il
d
−. 05
D ev
ia ti
o n
s fr
o m
p ar
ti ci
p an
ts ’
m ea
n d
el ay
s, co
m p
ar ed
w it
h “o
th er
” (i
.e .,
n o
p u
n is
h m
en t
o r
re as
o n
in g)
.C o
u ld
n o
t re
p li
ca te
G er
sh o
ff
R ea
so n
in g,
w h
et h
er u
se d
w it
h p
h ys
ic al
o r
n o
n p
h ys
ic al
p u
n is
h m
en t
o r
n o
t
D el
ay u
n ti
l d
is o
b ed
ie n
ce re
cu rr
en ce
.0 1
S am
e
N o
n p
h ys
ic al
p u
n is
h m
en t
(t im
e- o
u t
o r
p ri
vi le
ge re
m o
va l)
, w
h et
h er
u se
d w
it h
re as
o n
in g
o r
P P
o r
n o
t
D el
ay u
n ti
l d
is o
b ed
ie n
ce re
cu rr
en ce
−. 02
S am
e
R ea
so n
in g
& P
P (n
o n
o n
p h
ys ic
al p
u n
is h
m en
t; co
n d
it io
n al
P P
co m
b in
ed th
is w
it h
R ea
so n
in g
& P
P &
n o
n p
h ys
ic al
p u
n is
h m
en t)
D el
ay u
n ti
l d
is o
b ed
ie n
ce re
cu rr
en ce
−. 02
S am
e
R ea
so n
in g
& n
o n
p h
ys ic
al p
u n
is h
m en
t &
P P
(p ar
t o
f co
n d
it io
n al
P P
)
D el
ay u
n ti
l d
is o
b ed
ie n
ce re
cu rr
en ce
.1 7
S am
e
R ea
so n
in g
al o
n e
(n o
p h
ys ic
al o
r n
o n
p h
ys ic
al p
u n
is h
m en
t)
D el
ay u
n ti
l d
is o
b ed
ie n
ce re
cu rr
en ce
.0 6
S am
e
N o
n p
h ys
ic al
p u
n is
h m
en t
al o
n e
(t im
e- o
u t
o r
p ri
vi le
ge re
m o
va l;
n o
re as
o n
in g
o r
P P
)
D el
ay u
n ti
l d
is o
b ed
ie n
ce re
cu rr
en ce
.0 2
S am
e
R ea
so n
in g
& n
o n
p h
ys ic
al p
u n
is h
m en
t (n
o P
P )
D el
ay u
n ti
l d
is o
b ed
ie n
ce re
cu rr
en ce
−. 08
S am
e
Physical Punishment vs. Alternative Tactics 7 R
it ch
ie (1
99 9)
3 B
,G (M
) V
o lu
n te
er s
fr o
m b
ir th
re co
rd s
90 S
p an
k (c
o n
d it
io n
al P
P )
Im m
ed ia
te re
d u
ct io
n in
p ro
b ab
il it
y o
f d
efi an
ce
.9 7
F ro
m im
m ed
ia te
ly p
ri o
r p
ro b
ab il
it y
o f
d efi
an ce
co m
p ar
ed to
im m
ed ia
te ly
su b
se q
u en
t p
ro b
ab il
it y
o f
d efi
an ce
. S
tu d
y n
o t
in G
er sh
o ff
R ea
so n
o r
o ff
er al
te rn
at iv
es D
ro p
in d
efi an
ce −.
02 S
am e
T h
re at
en o
r ve
rb al
p o
w er
as se
rt io
n D
ro p
in d
efi an
ce .0
8 S
am e
P ri
vi le
ge re
m o
va l
D ro
p in
d efi
an ce
−. 33
S am
e T
im e-
o u
t D
ro p
in d
efi an
ce .6
0 S
am e
N o
re sp
o n
se (i
gn o
re )
D ro
p in
d efi
an ce
.0 2
S am
e P
h ys
ic al
p o
w er
as se
rt io
n D
ro p
in d
efi an
ce .4
5 S
am e
S p
an k
(c u
st o
m ar
y P
P )
Im m
ed ia
te re
d u
ct io
n in
“p h
ys ic
al ”
o r
p as
si ve
n o
n co
m p
li an
ce
.0 7
F ro
m im
m ed
ia te
ly p
ri o
r p
ro b
ab il
it y
o f
tw o
n o
n co
m p
li an
ce ty
p es
co m
p ar
ed to
th ei
r im
m ed
ia te
ly su
b se
q u
en t
p ro
b ab
il it
y R
ea so
n o
r o
ff er
al te
rn at
iv es
R ed
u ct
io n
in n
o n
co m
p li
an ce
.1 8
S am
e
T h
re at
en o
r ve
rb al
p o
w er
as se
rt io
n R
ed u
ct io
n in
n o
n co
m p
li an
ce .0
2 S
am e
P ri
vi le
ge re
m o
va l
R ed
u ct
io n
in n
o n
co m
p li
an ce
.2 4
S am
e
T im
e- o
u t
R ed
u ct
io n
in n
o n
co m
p li
an ce
.1 6
S am
e
N o
re sp
o n
se (i
gn o
re )
R ed
u ct
io n
in n
o n
co m
p li
an ce
.3 3
S am
e
P h
ys ic
al p
o w
er as
se rt
io n
R ed
u ct
io n
in n
o n
co m
p li
an ce
.2 0
S am
e
C o
rr el
at io
n al
se q
u en
ti al
an al
ys es
C h
ap m
an an
d Z
ah n
-W ax
le r
(1 98
2) 10
–2 9
m o
s. ,B
, G
(M )
V o
lu n
te er
s 24
P h
ys ic
al co
er ci
o n
(P P
o r
re st
ra in
t) w
it h
o u
t re
as o
n in
g (c
u st
o m
ar y
P P
)
Im m
ed ia
te co
m p
li an
ce .0
9 C
o m
p ar
ed to
th e
o ve
ra ll
co m
p li
an ce
ra te
.C o
u ld
n o
t re
p li
ca te
G er
sh o
ff
P h
ys ic
al co
er ci
o n
& re
as o
n in
g (c
o n
d it
io n
al P
P )
Im m
ed ia
te co
m p
li an
ce .0
2 C
o m
p ar
ed to
th e
o ve
ra ll
co m
p li
an ce
ra te
R ea
so n
in g
(w it
h o
r w
it h
o u
t ve
rb al
p ro
h ib
it io
n )
Im m
ed ia
te co
m p
li an
ce −.
22 S
am e
V er
b al
p ro
h ib
it io
n Im
m ed
ia te
co m
p li
an ce
−. 15
S am
e L
o ve
w it
h d
ra w
al (i
n cl
u d
in g
ig n
o ri
n g
an d
ti m
e- o
u t)
p lu
s an
y o
f ab
o ve
ta ct
ic s
Im m
ed ia
te co
m p
li an
ce .4
0 S
am e
M in
to n
,K ag
an ,a
n d
L ev
in e
(1 97
1) 27
m o
s. B
,G (M
) V
o lu
n te
er s
70 P
P as
p ro
p o
rt io
n o
f o
b se
rv ed
m is
b eh
av io
r (p
re d
o m
in an
t P
P )
D is
o b
ed ie
n ce
re q
u ir
in g
m at
er n
al re
p ri
m an
d
−. 55
A ve
ra ge
d co
rr el
at io
n s
fo r
b o
ys an
d gi
rl s.
S am
e as
G er
sh o
ff
8 Larzelere and Kuhn T
ab le
I. C
o n
ti n
u ed
B as
is o
f ef
fe ct
si ze
& A
ge ,G
en d
er D
is ci
p li
n e
E ff
ec t
d is
cr ep
an ci
es fr
o m
S tu
d y
(P ar
en t)
a S
am p
le N
ta ct
ic b
O u
tc o
m e
si ze
c (d
) G
er sh
o ff
(2 00
2)
E xp
la n
at io
n s
o f
re p
ri m
an d
s as
p ro
p o
rt io
n o
f m
is b
eh av
io r
O b
se rv
ed d
is o
b ed
ie n
ce .4
4 A
ve ra
ge d
co rr
el at
io n
s fo
r b
o ys
an d
gi rl
s
A n
ti so
ci al
b eh
av io
r S
ta ti
st ic
al ly
co n
tr o
ll ed
lo n
gi tu
d in
al st
u d
ie s
L ar
ze le
re an
d S
m it
h (2
00 0)
6– 9
at T
1, 8–
11 at
T 2,
B ,G
(M )
N at
io n
al sa
m p
le o
f yo
u n
g m
o th
er s
78 5
F re
q u
en cy
sp an
k ed
in p
as t
w ee
k (c
u st
o m
ar y
P P
)
A n
ti so
ci al
b eh
av io
r 2
ye ar
s la
te r
−. 23
M ea
n an
ti so
ci al
fo r
1 o
r m
o re
ti m
es p
er w
ee k
vs .n
o u
se o
f th
e d
is ci
p li
n ar
y ta
ct ic
,c o
n tr
o ll
in g
fo r
ex te
rn al
iz in
g p
ro b
le m
s at
T im
e 1,
fi ve
o th
er va
ri ab
le s
an d
6 in
te ra
ct io
n s
o f
th es
e va
ri ab
le s
w it
h th
e d
is ci
p li
n ar
y ta
ct ic
. U
n p
u b
li sh
ed ,n
o t
in G
er sh
o ff
78 5
F re
q u
en cy
p ri
vi le
ge s
re m
o ve
d in
p as
t w
ee k
A n
ti so
ci al
b eh
av io
r la
te r
−. 21
S am
e
78 5
F re
q u
en cy
gr o
u n
d ed
in p
as t
w ee
k A
n ti
so ci
al b
eh av
io r
la te
r −.
20 S
am e
77 1
F re
q u
en cy
al lo
w an
ce re
m o
ve d
in p
as t
w ee
k
A n
ti so
ci al
b eh
av io
r la
te r
−. 10
S am
e
78 5
F re
q u
en cy
se n
t to
ro o
m in
p as
t w
ee k
A n
ti so
ci al
b eh
av io
r la
te r
−. 18
S am
e
L ar
ze le
re et
al .
(1 99
8) 2–
3 at
T 1,
4 at
T 2;
B ,G
(M )
V o
lu n
te er
s 38
S la
p h
an d
o r
sp an
k (P
P )
w it
h o
u t
re as
o n
in g,
as p
ro p
o rt
io n
o f
m is
b eh
av io
r in
ci d
en ts
(a ve
ra ge
d w
it h
R ea
so n
in g
& P
P fo
r p
re d
o m
in an
t P
P )
D is
ru p
ti ve
b eh
av io
r 20
m o
n th
s la
te r
.4 1
M ea
n p
ar ti
al co
rr el
at io
n o
f p
ro p
o rt
io n
al u
sa ge
w it
h su
b se
q u
en t
d is
ru p
ti ve
b eh
av io
r, co
n tr
o ll
in g
fo r
in it
ia l
d is
ru p
ti ve
b eh
av io
r. N
o t
in G
er sh
o ff
(m u
lt ip
le re
p o
rt s
fr o
m sa
m e
st u
d y)
R ea
so n
in g
& P
P (p
ro p
o rt
io n
al u
se ;
p ar
t o
f p
re d
o m
in an
t P
P )
D is
ru p
ti ve
b eh
av io
r 20
m o
n th
s la
te r
−. 32
S am
e
R ea
so n
in g
w it
h o
u t
P P
o r
n o
n p
h ys
ic al
p u
n is
h m
en t
D is
ru p
ti ve
b eh
av io
r la
te r
−. 80
S am
e
N o
n p
h ys
ic al
p u
n is
h m
en t
(t im
e- o
u t
o r
p ri
vi le
ge re
m o
va l)
D is
ru p
ti ve
b eh
av io
r la
te r
.2 0
S am
e
R ea
so n
in g
& n
o n
p h
ys ic
al p
u n
is h
m en
t
D is
ru p
ti ve
b eh
av io
r la
te r
.1 0
S am
e
Physical Punishment vs. Alternative Tactics 9
W it
h in
-s u
b je
ct se
q u
en ti
al an
al ys
es L
ar ze
le re
et al
. (1
99 6)
2– 3
B ,G
(M )
V o
lu n
te er
s 38
S la
p h
an d
o r
sp an
k (P
P ),
w h
et h
er re
as o
n in
g w
as al
so u
se d
o r
n o
t (c
u st
o m
ar y
P P
, co
n tr
as te
d w
it h
n ex
t tw
o ta
ct ic
s)
D el
ay u
n ti
l n
ex t
re cu
rr en
ce o
f fi
gh ti
n g,
co m
p ar
ed to
ty p
ic al
d el
ay s
fo r
th at
ch il
d
.0 8
D ev
ia ti
o n
s fr
o m
p ar
ti ci
p an
ts ’
m ea
n d
el ay
s, co
m p
ar ed
w it
h “o
th er
” (i
.e .,
n o
p u
n is
h m
en t
o r
re as
o n
in g)
.G er
sh o
ff d
id n
o t
in cl
u d
e th
is o
u tc
o m
e
R ea
so n
in g,
w h
et h
er u
se d
w it
h p
h ys
ic al
o r
n o
n p
h ys
ic al
p u
n is
h m
en t
o r
n o
t
D el
ay u
n ti
l fi
gh ti
n g
re cu
rr en
ce .0
1 S
am e
N o
n p
h ys
ic al
p u
n is
h m
en t
(t im
e- o
u t
o r
p ri
vi le
ge re
m o
va l)
, w
h et
h er
u se
d w
it h
re as
o n
in g
o r
P P
o r
n o
t
D el
ay u
n ti
l fi
gh ti
n g
re cu
rr en
ce .3
6 S
am e
R ea
so n
in g
& P
P (n
o n
o n
p h
ys ic
al p
u n
is h
m en
t; (c
o n
d it
io n
al P
P co
m b
in ed
th is
w it
h R
ea so
n in
g &
P P
& n
o n
p h
ys ic
al p
u n
is h
m en
t)
D el
ay u
n ti
l fi
gh ti
n g
re cu
rr en
ce (d
ev ia
ti o
n )
.0 7
S am
e
R ea
so n
in g
& n
o n
p h
ys ic
al p
u n
is h
m en
t &
P P
(p ar
t o
f co
n d
it io
n al
P P
)
D el
ay u
n ti
l fi
gh ti
n g
re cu
rr en
ce .8
1 S
am e
R ea
so n
in g
al o
n e
(n o
p h
ys ic
al o
r n
o n
p h
ys ic
al p
u n
is h
m en
t)
D el
ay u
n ti
l fi
gh ti
n g
re cu
rr en
ce −.
09 S
am e
N o
n p
h ys
ic al
p u
n is
h m
en t
al o
n e
(t im
e- o
u t
o r
p ri
vi le
ge re
m o
va l;
n o
re as
o n
in g
o r
P P
)
D el
ay u
n ti
l fi
gh ti
n g
re cu
rr en
ce .2
6 S
am e
R ea
so n
in g
& n
o n
p h
ys ic
al p
u n
is h
m en
t (n
o P
P )
D el
ay u
n ti
l fi
gh ti
n g
re cu
rr en
ce .6
3 S
am e
U n
co n
tr o
ll ed
lo n
gi tu
d in
al st
u d
ie s
M cC
le ll
an d
an d
P il
o n
(1 98
3) 5
at T
1, 31
at T
2; B
,G (M
) K
in d
er ga
rt en
sa m
p le
78 E
xt en
t o
f P
P ,
co m
b in
in g
fr eq
u en
cy an
d se
ve ri
ty (s
ev er
e P
P )
N ee
d fo
r P
o w
er 26
ye ar
s la
te r
.4 2
M ea
n o
f co
rr el
at io
n s
w it
h N
ee d
fo r
P o
w er
fo r
m al
es an
d fe
m al
es .S
tu d
y n
o t
in G
er sh
o ff
R ea
so n
in g
N ee
d fo
r P
o w
er .0
0 N
o n
-s ig
n ifi
ca n
t r
P ri
vi le
ge re
m o
va l
N ee
d fo
r P
o w
er .0
0 N
o n
-s ig
n ifi
ca n
t r
10 Larzelere and Kuhn T
ab le
I. C
o n
ti n
u ed
B as
is o
f ef
fe ct
si ze
& A
ge ,G
en d
er D
is ci
p li
n e
E ff
ec t
d is
cr ep
an ci
es fr
o m
S tu
d y
(P ar
en t)
a S
am p
le N
ta ct
ic b
O u
tc o
m e
si ze
c (d
) G
er sh
o ff
(2 00
2)
L o
ve w
it h
d ra
w al
N ee
d fo
r P
o w
er .0
0 N
o n
-s ig
n ifi
ca n
t r
S ea
rs (1
96 1)
5 at
T 1,
12 at
T 2;
B ,G
(M )
K in
d er
ga rt
en sa
m p
le 16
0 E
xt en
t o
f P
P ,
co m
b in
in g
se ve
ri ty
an d
fr eq
u en
cy (s
ev er
e P
P )
A n
ti so
ci al
ag gr
es si
o n
7 ye
ar s
la te
r .1
4 C
o rr
el at
io n
s w
it h
an ti
so ci
al ag
gr es
si o
n 7
ye ar
s la
te r.
G er
sh o
ff av
er ag
ed 6
co rr
el at
io n
s (c
ro ss
-s ec
ti o
n al
an d
w it
h p
ro so
ci al
,a m
b ig
u o
u s,
an d
an ti
so ci
al ag
gr es
si o
n )
P ri
vi le
ge re
m o
va l
A n
ti so
ci al
ag gr
es si
o n
−. 12
C o
rr el
at io
n s
w it
h la
te r
an ti
so ci
al ag
gr es
si o
n L
o ve
w it
h d
ra w
al A
n ti
so ci
al ag
gr es
si o
n .1
1 S
am e
Y ar
ro w
et al
.( 19
68 )
4 B
,G (M
) N
u rs
er y
sc h
o o
l sa
m p
le 58
U se
o f
p h
ys ic
al p
u n
is h
m en
t fo
r vi
gn et
te s
ab o
u t
ex tr
em e
d is
o b
ed ie
n ce
(c o
n d
it io
n al
P P
)
T ea
ch er
-r at
ed ag
gr es
si o
n in
n u
rs er
y sc
h o
o l
2 m
o n
th s
la te
r
.3 8
C o
rr el
at io
n .G
er sh
o ff
u se
d th
e co
rr el
at io
n o
f se
ve ri
ty o
f al
l p
u n
is h
m en
t fo
r ag
gr es
si o
n w
it h
co n
cu rr
en t
ag gr
es si
o n
to w
ar d
p ar
en ts
(m o
th er
-r ep
o rt
)
U se
o f
re as
o n
in g
fr o
m vi
gn et
te s
L at
er sc
h o
o l
ag gr
es si
o n
−. 28
C o
rr el
at io
n
U se
o f
sc o
ld in
g fr
o m
vi gn
et te
s L
at er
sc h
o o
l ag
gr es
si o
n −.
24 S
am e
U se
o f
p ri
vi le
ge re
m o
va lf
ro m
vi gn
et te
s
L at
er sc
h o
o l
ag gr
es si
o n
.3 8
S am
e
U se
o f
is o
la ti
o n
fr o
m vi
gn et
te s
L at
er sc
h o
o l
ag gr
es si
o n
−. 18
S am
e
U se
o f
d iv
er ti
n g
fr o
m vi
gn et
te s
L at
er sc
h o
o l
ag gr
es si
o n
−. 47
S am
e
U se
o f
lo ve
w it
h d
ra w
al fr
o m
vi gn
et te
s
L at
er sc
h o
o l
ag gr
es si
o n
−. 24
S am
e
R et
ro sp
ec ti
ve st
u d
ie s
W at
so n
(1 98
9) 0–
5 at
T 1;
17 at
T 2;
B ,G
(M ,
F )
N at
io n
al M
er it
S ch
o la
rs h
ip fi
n al
is ts
& av
er ag
e te
st -t
ak er
s
25 00
P ar
en t-
re p
o rt
ed sp
an k
in g
an d
p o
ss ib
ly ti
m e
o u
t b
ef o
re ag
e 6,
(c u
st o
m ar
y P
P )
Y o
u th
-r ep
o rt
ed h
o st
il it
y &
m il
d er
(“ o
b lo
q u
ia l”
) p
ro b
le m
s
−. 09
C o
rr el
at io
n s
w it
h h
o st
il it
y an
d o
b lo
q u
ia l
p ro
b le
m s,
u si
n g
.0 0
fo r
n o
n -s
ig n
ifi ca
n t
rs .G
er sh
o ff
u se
d o
n ly
th e
si gn
ifi ca
n t
r w
it h
o n
e h
o st
il it
y m
ea su
re
P ri
vi le
ge re
m o
va la
n d
as si
gn in
g ex
tr a
d u
ti es
b ef
o re
ag e
6
H o
st il
it y
an d
m il
d er
b eh
av io
r p
ro b
le m
s −.
13 C
o rr
el at
io n
s w
it h
h o
st il
it y
& o
b lo
q u
ia l
p ro
b le
m s,
u si
n g
.0 0
fo r
n o
n -s
ig n
ifi ca
n t
rs U
n co
n tr
o ll
ed cr
o ss
-s ec
ti o
n al
st u
d ie
s S
tr au
s an
d M
o u
ra d
ia n
(1 99
8) 2–
14 B
,G (M
) R
an d
o m
sa m
p le
o f
tw o
co u
n ti
es 74
4 H
o w
o ft
en “s
p an
k ed
, sl
ap p
ed o
r h
it ”
th e
ch il
d d
u ri
n g
th e
p as
t 6
m o
n th
s, co
n tr
o ll
in g
fo r
se ve
re o
u t-
o f-
co n
tr o
l P
P (c
o n
d it
io n
al P
P )
A n
ti so
ci al
an d
im p
u ls
iv e
b eh
av io
r −.
14 F
-v al
u es
fo r
P P
,c o
n tr
o ll
in g
fo r
se ve
re P
P ,3
o th
er d
is ci
p li
n ar
y ta
ct ic
s, 4
o th
er va
ri ab
le s,
an d
th ei
r in
te ra
ct io
n s
w it
h P
P .G
er sh
o ff
p ro
b ab
ly u
se d
gr ap
h ed
m ea
n an
ti so
ci al
sc o
re s,
w h
ic h
co u
ld n
o t
b e
co m
p ar
ed w
it h
al te
rn at
iv e
ta ct
ic s
Physical Punishment vs. Alternative Tactics 11
P er
ce n
ta ge
o f
sp an
k in
gs in
w h
ic h
m o
th er
s sa
id th
ey “l
o st
it ”
d u
e to
an ge
r (s
ev er
e P
P )
A n
ti so
ci al
an d
im p
u ls
iv e
b eh
av io
r −.
28 F
-v al
u es
fo r
se ve
re P
P ,w
it h
ab o
ve co
n tr
o ls
H o
w o
ft en
th ey
u se
d d
is ci
p li
n ar
y re
as o
n in
g, p
ri vi
le ge
re m
o va
l, an
d ti
m e-
o u
t d
u ri
n g
p as
t 6
m o
n th
s
A n
ti so
ci al
an d
im p
u ls
iv e
b eh
av io
r −.
39 F
-v al
u es
fo r
th es
e al
te rn
at iv
e d
is ci
p li
n ar
y ta
ct ic
s, w
it h
ab o
ve co
n tr
o ls
S u
b st
an ce
ab u
se R
et ro
sp ec
ti ve
st u
d ie
s T
en n
an t
et al
.( 19
75 )
0– 14
at T
1, M
= 23
at T
2; B
(M ,F
)
U S
A rm
y so
ld ie
rs 50
44 S
p an
k in
g (c
u st
o m
ar y
P P
) F
re q
u en
t u
se o
f h
as h
is h
,a lc
o h
o l,
am p
h et
am in
es ,a
n d
o p
ia te
s
.2 4
P er
ce n
ta ge
o f
m o
st an
d le
as t
fr eq
u en
t u
se rs
re p
o rt
in g
b ei
n g
sp an
k ed
. A
ve ra
ge d
ac ro
ss 4
su b
st an
ce s.
N o
t am
o n
g G
er sh
o ff
’s 11
o u
tc o
m es
N o
n -c
o n
ta ct
p u
n is
h m
en t
F re
q u
en t
u se
o f
h as
h is
h ,a
lc o
h o
l, am
p h
et am
in es
,a n
d o
p ia
te s
−. 08
P er
ce n
ta ge
o f
m o
st vs
.l ea
st fr
eq u
en t
u se
rs re
p o
rt in
g re
ce iv
in g
th is
p u
n is
h m
en t.
A ve
ra ge
d ac
ro ss
4 su
b st
an ce
s W
at so
n (1
98 9)
0– 5
at T
im e
1; 17
at T
im e
2. B
, G
(B )
N at
io n
al M
er it
S ch
o la
rs h
ip fi
n al
is ts
& av
er ag
e te
st -t
ak er
s
25 00
P ar
en ta
lr ep
o rt
o f
sp an
k in
g an
d p
o ss
ib ly
ti m
e o
u t
b ef
o re
ag e
6 (c
u st
o m
ar y
P P
)
Y o
u th
-r ep
o rt
ed al
co h
o lu
sa ge
−. 02
C o
rr el
at io
n .O
u tc
o m
e n
o t
in G
er sh
o ff
W it
h d
ra w
al o
f p
ri vi
le ge
s an
d as
si gn
in g
ex tr
a d
u ti
es b
ef o
re ag
e 6
Y o
u th
-r ep
o rt
ed al
co h
o lu
sa ge
−. 10
C o
rr el
at io
n
C o
n sc
ie n
ce &
re si
st an
ce to
te m
p ta
ti o
n U
n co
n tr
o ll
ed lo
n gi
tu d
in al
st u
d ie
s G
ri n
d er
(1 96
2) 5–
6 at
T 1,
11 –1
2 at
T 2;
B ,G
(M )
K in
d er
ga rt
en sa
m p
le 14
0 E
xt en
t o
f P
P ,
co m
b in
in g
se ve
ri ty
an d
fr eq
u en
cy (s
ev er
e P
P )
R es
is ts
te m
p ta
ti o
n in
fo rb
id d
en -t
o y
la b
te st
6 ye
ar s
la te
r
−. 24
T h
e p
er ce
n ta
ge o
f ch
il d
re n
w h
o re
si st
ed te
m p
ta ti
o n
fo r
h ig
h vs
.l o
w u
se o
f th
e d
is ci
p li
n ar
y ta
ct ic
.C o
u ld
n o
t re
p li
ca te
G er
sh o
ff ’s
b en
efi ci
al ef
fe ct
si ze
P ri
vi le
ge re
m o
va l
R es
is ts
te m
p ta
ti o
n .1
0 S
am e
Is o
la ti
o n
R es
is ts
te m
p ta
ti o
n −.
01 S
am e
L o
ve w
it h
d ra
w al
R es
is ts
te m
p ta
ti o
n .3
7 S
am e
U n
co n
tr o
ll ed
cr o
ss -s
ec ti
o n
al st
u d
ie s
A ro
n fr
ee d
(1 96
1) 12
B ,G
(M )
S ix
th -g
ra d
e sa
m p
le 12
0 P
ri m
ar il
y se
n si
ti za
ti o
n (P
P &
u n
co n
tr o
ll ed
ve rb
al as
sa u
lt s)
vs .
p ri
m ar
il y
in d
u ct
io n
(l o
ve w
it h
d ra
w al
, en
co u
ra gi
n g
re sp
o n
si b
il it
y, an
d ex
p la
n at
io n
s; b
as ed
o n
re sp
o n
se s
to vi
gn et
te s;
p re
d o
m in
an t
P P
)
In te
rn al
an d
ex te
rn al
m o
ti va
ti o
n s
fo r
m o
ra lc
o rr
ec ti
o n
s in
p ro
je ct
iv e
st o
ry co
m p
le ti
o n
s
−. 16
2 ×
2 co
n ti
n ge
n cy
ta b
le s
w it
h 6
in te
rn al
an d
6 ex
te rn
al m
o ra
lm o
ti va
ti o
n s,
co u
n ti
n g
9 n
o n
-s ig
n ifi
ca n
t as
so ci
at io
n s
as .0
0. G
er sh
o ff
u se
d o
n ly
th e
si gn
ifi ca
n t
as so
ci at
io n
w it
h re
p ar
at io
n s
12 Larzelere and Kuhn
T ab
le I.
C o
n ti
n u
ed
B as
is o
f ef
fe ct
si ze
& A
ge ,G
en d
er D
is ci
p li
n e
E ff
ec t
d is
cr ep
an ci
es fr
o m
S tu
d y
(P ar
en t)
a S
am p
le N
ta ct
ic b
O u
tc o
m e
si ze
c (d
) G
er sh
o ff
(2 00
2)
B u
rt o
n et
al .(
19 61
) 4
B ,G
(M )
P ri
va te
n u
rs er
y sc
h o
o l
sa m
p le
77 In
te rv
ie w
er ra
ti n
g o
f P
P as
u su
al d
is ci
p li
n e
te ch
n iq
u e
(p re
d o
m in
an t
P P
)
R es
is ts
te m
p ta
ti o
n (l
ab te
st )
.3 5
F ro
m 2
× 2
co n
ti n
ge n
cy ta
b le
.G er
sh o
ff d
id n
o t
in cl
u d
e th
is o
u tc
o m
e
R ea
so n
in g
ra te
d as
u su
al te
ch n
iq u
e R
es is
ts te
m p
ta ti
o n
−. 08
2 ×
2 co
n ti
n ge
n cy
ta b
le
S co
ld in
g ra
te d
as u
su al
te ch
n iq
u e
R es
is ts
te m
p ta
ti o
n .2
5 S
am e
P ri
vi le
ge re
m o
va l
ra te
d as
u su
al te
ch n
iq u
e
R es
is ts
te m
p ta
ti o
n −.
63 S
am e
Is o
la ti
o n
ra te
d as
u su
al te
ch n
iq u
e R
es is
ts te
m p
ta ti
o n
−. 12
S am
e
In te
rv ie
w er
ra ti
n g
o f
fr eq
u en
cy o
f sp
an k
in g,
sl ap
p in
g, an
d sh
ak in
g (s
ev er
e P
P )
R es
is ts
te m
p ta
ti o
n .0
0 T
ab le
n o
t gi
ve n
,e st
im at
ed at
.0 0
d u
e to
n o
n -s
ig n
ifi ca
n ce
R at
ed fr
eq u
en cy
o f
re as
o n
in g
R es
is ts
te m
p ta
ti o
n −.
61 2
× 2
co n
ti n
ge n
cy ta
b le
R at
ed fr
eq u
en cy
o f
sc o
ld in
g R
es is
ts te
m p
ta ti
o n
.0 0
N o
n -s
ig n
ifi ca
n t
as so
ci at
io n
R at
ed fr
eq u
en cy
o f
p ri
vi le
ge re
m o
va l
R es
is ts
te m
p ta
ti o
n −.
34 2
× 2
co n
ti n
ge n
cy ta
b le
.
R at
ed fr
eq u
en cy
o f
is o
la ti
o n
R es
is ts
te m
p ta
ti o
n .0
0 N
o n
-s ig
n ifi
ca n
t as
so ci
at io
n
R at
ed fr
eq u
en cy
o f
lo ve
w it
h d
ra w
al R
es is
ts te
m p
ta ti
o n
.3 3
2 ×
2 co
n ti
n ge
n cy
ta b
le
In te
rv ie
w er
ra ti
n g
o f
fr eq
u en
cy o
f sp
an k
in g,
sl ap
p in
g, an
d sh
ak in
g (s
ev er
e P
P )
C o
n sc
ie n
ce (i
n it
ia l
ch il
d ac
ti o
n s
af te
r w
ro n
gd o
in g)
−. 62
G er
sh o
ff u
se d
o n
ly th
is co
rr el
at io
n fr
o m
th e
st u
d y
R at
ed fr
eq u
en cy
o f
re as
o n
in g
C o
n sc
ie n
ce .3
6 C
o rr
el at
io n
R at
ed fr
eq u
en cy
o f
sc o
ld in
g C
o n
sc ie
n ce
.0 0
N o
n -s
ig n
ifi ca
n t
r
R at
ed fr
eq u
en cy
o f
p ri
vi le
ge re
m o
va l
C o
n sc
ie n
ce −.
32 C
o rr
el at
io n
R at
ed fr
eq u
en cy
o f
is o
la ti
o n
C o
n sc
ie n
ce .3
0 C
o rr
el at
io n
R at
ed fr
eq u
en cy
o f
lo ve
w it
h d
ra w
al C
o n
sc ie
n ce
.0 0
N o
n -s
ig n
ifi ca
n t
r
Physical Punishment vs. Alternative Tactics 13 L
yt to
n (1
97 7)
2 B
(M ,F
) V
o lu
n te
er s
90 M
at er
n al
an d
p at
er n
al fr
eq u
en ci
es o
f P
P (c
u st
o m
ar y
P P
)
C o
m p
li an
ce (t
w o
m ea
su re
s) an
d co
n sc
ie n
ce (1
)
−. 04
A ve
ra ge
o f
b et
as o
f 2
p ar
en ts
fo r
3 o
u tc
o m
es ,a
ss u
m in
g .0
0 fo
r n
o n
-s ig
n ifi
ca n
t p
re d
ic to
rs .G
er sh
o ff
o n
ly u
se d
th e
o n
e si
gn ifi
ca n
t co
rr el
at io
n R
at in
g o
f m
o th
er s’
in d
u ct
io n
(e .g
., ex
p la
n at
io n
o f
o rd
er s)
C o
m p
li an
ce an
d co
n sc
ie n
ce .0
6 N
o n
-s ig
n ifi
ca n
t rs
,b u
t m
o st
ly in
th e
in d
ic at
ed d
ir ec
ti o
n ,a
cc o
rd in
g to
th e
te xt
R at
in g
o f
m o
th er
s’ ve
rb al
p sy
ch o
lo gi
ca l
p u
n is
h m
en t
(c ri
ti ci
sm ,
w it
h d
ra w
al o
f lo
ve )
C o
m p
li an
ce an
d co
n sc
ie n
ce −.
11 A
ve ra
ge o
f b
et as
fo r
th re
e o
u tc
o m
es ,
as su
m in
g .0
0 fo
r n
o n
-s ig
n ifi
ca n
t p
re d
ic to
rs
M at
er n
al fr
eq u
en cy
o f
lo ve
w it
h d
ra w
al C
o m
p li
an ce
an d
co n
sc ie
n ce
−. 06
N o
n -s
ig n
ifi ca
n t
rs ,b
u t
m o
st ly
in th
e in
d ic
at ed
d ir
ec ti
o n
,a cc
o rd
in g
to th
e te
xt S
ea rs
et al
.( 19
57 )
5 B
,G (M
) K
in d
er ga
rt en
sa m
p le
16 0
E xt
en t
o f
P P
, co
m b
in in
g se
ve ri
ty an
d fr
eq u
en cy
(s ev
er e
P P
)
C o
n sc
ie n
ce −.
41 C
o rr
el at
io n
s o
f h
ig h
vs .l
o w
u se
o f
ea ch
d is
ci p
li n
ar y
ta ct
ic w
it h
co n
sc ie
n ce
, si
m il
ar to
G er
sh o
ff
R ea
so n
in g
C o
n sc
ie n
ce .3
7 S
am e
P ri
vi le
ge re
m o
va l
C o
n sc
ie n
ce −.
14 S
am e
Is o
la ti
o n
C o
n sc
ie n
ce .0
0 N
o n
-s ig
n ifi
ca n
t r
L o
ve w
it h
d ra
w al
C o
n sc
ie n
ce .1
8 C
o rr
el at
io n
s o
f h
ig h
vs .l
o w
lo ve
w it
h d
ra w
al &
co n
sc ie
n ce
Y ar
ro w
et al
.( 19
68 )
4 B
,G (M
) N
u rs
er y
sc h
o o
l sa
m p
le 86
R ep
o rt
ed u
se o
f p
h ys
ic al
p u
n is
h m
en t
fr o
m vi
gn et
te s
ab o
u t
ex tr
em e
d is
o b
ed ie
n ce
(c o
n d
it io
n al
P P
)
C o
n sc
ie n
ce (m
at er
n al
re p
o rt
) −.
02 C
o rr
el at
io n
.T h
is o
u tc
o m
e n
o t
in G
er sh
o ff
U se
o f
re as
o n
in g
fr o
m vi
gn et
te s
C o
n sc
ie n
ce .2
2 C
o rr
el at
io n
U se
o f
sc o
ld in
g fr
o m
vi gn
et te
s C
o n
sc ie
n ce
−. 18
S am
e
U se
o f
p ri
vi le
ge re
m o
va lf
ro m
vi gn
et te
s
C o
n sc
ie n
ce .2
2 S
am e
U se
o f
is o
la ti
o n
fr o
m vi
gn et
te s
C o
n sc
ie n
ce −.
30 S
am e
U se
o f
d iv
er ti
n g
fr o
m vi
gn et
te s
C o
n sc
ie n
ce .3
2 S
am e
U se
o f
lo ve
w it
h d
ra w
al fr
o m
vi gn
et te
s
C o
n sc
ie n
ce −.
04 S
am e
14 Larzelere and Kuhn
T ab
le I.
C o
n ti
n u
ed
B as
is o
f ef
fe ct
si ze
& A
ge ,G
en d
er D
is ci
p li
n e
E ff
ec t
d is
cr ep
an ci
es fr
o m
S tu
d y
(P ar
en t)
a S
am p
le N
ta ct
ic b
O u
tc o
m e
si ze
c (d
) G
er sh
o ff
(2 00
2)
P ro
so ci
al b
eh av
io r
U n
co n
tr o
ll ed
lo n
gi tu
d in
al st
u d
ie s
Z ah
n -W
ax le
r, R
ad k
e- Y
ar ro
w ,
an d
K in
g (1
97 9)
15 –2
4 m
o n
th s
B ,
G (M
) V
o lu
n te
er s
16 P
ro p
o rt
io n
al u
se o
f P
P in
em o
ti o
n al
ly ch
ar ge
d m
o th
er –c
h il
d in
te ra
ct io
n s
(p re
d o
m in
an t
P P
)
R ep
ar at
io n
s an
d al
tr u
is m
4. 5
m o
n th
s la
te r
−. 58
A ve
ra ge
co rr
el at
io n
s fo
r ti
m e-
la g
d at
a fo
r re
p ar
at io
n s
an d
fo r
al tr
u is
m .
G er
sh o
ff p
ro b
ab ly
u se
d th
es e
p lu
s co
n te
m p
o ra
n eo
u s
rs
E xp
la n
at io
n s
w it
h af
fe ct
(p ro
p o
rt io
n al
u se
)
R ep
ar at
io n
s an
d al
tr u
is m
la te
r 1.
25 S
am e
P o
si ti
ve su
gg es
ti o
n s
(p ro
p o
rt io
n al
u se
) R
ep ar
at io
n s
an d
al tr
u is
m la
te r
.5 6
S am
e
E xp
la n
at io
n s
w it
h n
eu tr
al af
fe ct
(p ro
p o
rt io
n al
u se
)
R ep
ar at
io n
s an
d al
tr u
is m
la te
r −.
10 S
am e
P h
ys ic
al re
st ra
in t
(p ro
p o
rt io
n al
u se
) R
ep ar
at io
n s
an d
al tr
u is
m la
te r
o n
.2 6
S am
e
U n
ex p
la in
ed ve
rb al
p ro
h ib
it io
n s
(p ro
p o
rt io
n al
u se
)
R ep
ar at
io n
s an
d al
tr u
is m
la te
r o
n −.
70 S
am e
Ig n
o ri
n g
(p ro
p o
rt io
n al
u se
) R
ep ar
at io
n s
an d
al tr
u is
m la
te r
o n
−. 34
S am
e
R et
ro sp
ec ti
ve st
u d
ie s
W at
so n
(1 98
9) 0–
5 at
T 1;
17 at
T 2;
B ,G
(M ,
F )
N at
io n
al M
er it
S ch
o la
rs h
ip fi
n al
is ts
& av
er ag
e te
st -t
ak er
s
25 00
P ar
en ta
lr ep
o rt
o f
sp an
k in
g an
d p
o ss
ib ly
ti m
e o
u t
b ef
o re
ag e
6 (c
u st
o m
ar y
P P
)
Y o
u th
-r ep
o rt
ed al
tr u
is m
.0 0
C o
rr el
at io
n ,u
si n
g .0
0 fo
r n
o n
-s ig
n ifi
ca n
t rs
.O u
tc o
m e
n o
t in
G er
sh o
ff
P ar
en ta
lr ep
o rt
o f
w it
h d
ra w
al o
f p
ri vi
le ge
s an
d as
si gn
in g
ex tr
a d
u ti
es b
ef o
re ag
e 6
Y o
u th
-r ep
o rt
ed al
tr u
is m
.0 0
N o
n -s
ig n
ifi ca
n t
rs
U n
co n
tr o
ll ed
cr o
ss -s
ec ti
o n
al st
u d
ie s
H al
l( 19
94 )
4– 5
B ,G
(M )
P re
sc h
o o
le rs
fr o
m h
ig h
in co
m e
fa m
il ie
s
41 F
re q
u en
cy o
f sp
an k
in g
o r
sl ap
p in
g, fr
o m
C o
n fl
ic t
T ac
ti cs
S ca
le (C
T S
) it
em s
(c u
st o
m ar
y P
P )
V er
b al
p o
si ti
ve s
o n
in te
rp er
so n
al p
ro b
le m
so lv
in g
ta sk
−. 18
C o
rr el
at io
n .G
er sh
o ff
u se
d o
n ly
th e
r w
it h
n o
n ve
rb al
n eg
at iv
es o
n a
si m
il ar
ta sk
,b u
t th
e eq
u iv
al en
t r
w as
n o
t re
p o
rt ed
fo r
re as
o n
in g
R ea
so n
in g
(% o
f m
ax im
u m
p o
ss ib
le sc
o re
o n
th re
e C
T S
it em
s)
V er
b al
p o
si ti
ve s
o n
in te
rp er
so n
al p
ro b
le m
so lv
in g
ta sk
.4 4
C o
rr el
at io
n
Physical Punishment vs. Alternative Tactics 15
S el
f- es
te em
R et
ro sp
ec ti
ve st
u d
ie s
L ar
ze le
re ,K
le in
, S
ch u
m m
,a n
d A
li b
ra n
d o
(1 98
9)
0– 12
at T
1, M
= 21
at T
2; B
,G (M
,F )
H o
m e
E co
n o
m ic
s co
ll eg
e st
u d
en ts
15 7
S p
an k
in g
fr eq
u en
cy (c
u st
o m
ar y
P P
) S
el f-
es te
em −.
10 C
o rr
el at
io n
in fu
ll re
p o
rt m
en ti
o n
ed in
ar ti
cl e’
s fo
o tn
o te
.O u
tc o
m e
n o
t in
G er
sh o
ff d
u e
to n
o n
-s ig
n ifi
ca n
t rs
A lt
er n
at iv
e p
u n
is h
m en
ts (t
im e-
o u
t, p
ri vi
le ge
re m
o va
l, re
st it
u ti
o n
) in
al l
th re
e ag
e gr
o u
p s
S el
f- es
te em
−. 08
C o
rr el
at io
n
W at
so n
(1 98
9) 0–
5 at
T 1;
17 at
T 2;
B ,G
(M ,
F )
N at
io n
al M
er it
S ch
o la
rs h
ip fi
n al
is ts
& av
er ag
e te
st -t
ak er
s
25 00
P ar
en ta
lr ep
o rt
o f
sp an
k in
g an
d p
o ss
ib ly
ti m
e o
u t
b ef
o re
ag e
6 (c
u st
o m
ar y
P P
)
Y o
u th
-r ep
o rt
ed n
eu ro
ti ci
sm ,
se lf
-a cc
ep ta
n ce
,a n
d se
n se
o f
w el
l- b
ei n
g
−. 03
C o
rr el
at io
n s
w it
h th
re e
o u
tc o
m es
,u si
n g
.0 0
fo r
n o
n -s
ig n
ifi ca
n t
rs .O
u tc
o m
e n
o t
in G
er sh
o ff
P ar
en ta
lr ep
o rt
o f
w it
h d
ra w
al o
f p
ri vi
le ge
s an
d as
si gn
in g
ex tr
a d
u ti
es b
ef o
re ag
e 6
N eu
ro ti
ci sm
, se
lf -a
cc ep
ta n
ce ,&
w el
l- b
ei n
g
.0 0
N o
n -s
ig n
ifi ca
n t
rs
U n
co n
tr o
ll ed
cr o
ss -s
ec ti
o n
al st
u d
ie s
C o
o p
er sm
it h
(1 96
7) 10
–1 2
B (M
) F
if th
-g ra
d er
s se
le ct
ed fo
r co
n si
st en
cy o
r in
co n
si st
en cy
o n
m ea
su re
s o
f se
lf -e
st ee
m
63 P
re d
o m
in an
t u
se o
f p
h ys
ic al
p u
n is
h m
en t
ra th
er th
an n
ex t
tw o
ta ct
ic s,
w h
en ru
le s
ar e
vi o
la te
d (p
re d
o m
in an
t P
P )
S el
f- es
te em
−. 42
P ro
p o
rt io
n s
o f
th e
h ig
h an
d lo
w se
lf -e
st ee
m gr
o u
p s
w it
h P
P as
p re
d o
m in
an t
d is
ci p
li n
ar y
m et
h o
d ,
co m
p ar
ed to
th e
n ex
t tw
o ,t
ac ti
cs si
m il
ar to
G er
sh o
ff
P re
d o
m in
an t
u se
o f
lo ve
w it
h d
ra w
al S
el f-
es te
em −.
54 P
ro p
o rt
io n
s o
f th
e ex
tr em
e se
lf -e
st ee
m gr
o u
p s
u si
n g
lo ve
w it
h d
ra w
al p
re d
o m
in an
tl y
P re
d o
m in
an t
u se
o f
m il
d er
m an
ag em
en t
ta ct
ic s
(“ re
st ra
in t,
d en
ia l,
is o
la ti
o n
”) m
o re
th an
ab o
ve tw
o ta
ct ic
s
S el
f- es
te em
.8 6
P ro
p o
rt io
n s
o f
th e
ex tr
em e
se lf
-e st
ee m
gr o
u p
s u
si n
g m
il d
er ta
ct ic
s p
re d
o m
in an
tl y
S tr
es s
d is
cu ss
io n
an d
re as
o n
in g
to o
b ta
in co
m p
li an
ce an
d co
o p
er at
io n
,r at
h er
th an
fo rc
e o
r au
to cr
at ic
m ea
n s
S el
f- es
te em
.8 7
P ro
p o
rt io
n s
o f
th e
ex tr
em e
se lf
-e st
ee m
gr o
u p
s th
at p
re fe
rr ed
ei th
er o
f th
es e
tw o
d is
ci p
li n
ar y
ap p
ro ac
h es
C o
m p
et en
cy U
n co
n tr
o ll
ed lo
n gi
tu d
in al
st u
d ie
s C
ro w
n e,
C o
n n
, M
ar lo
w e,
an d
E d
w ar
d s
(1 96
9)
5 at
T 1,
18 at
T 2;
B ,G
(M ,F
) K
in d
er ga
rt en
ch il
d re
n 83
F re
q u
en cy
an d
se ve
ri ty
o f
sp an
k in
g b
y ea
ch p
ar en
t (s
ev er
e P
P )
A m
b it
io u
s, ye
t re
al is
ti c
as p
ir at
io n
s an
d u
n u
su al
go al
sh if
ts 13
ye ar
s la
te r
−. 00
M ea
n o
f 15
as so
ci at
io n
s fo
r th
re e
m ea
su re
s o
f p
h ys
ic al
p u
n is
h m
en t
an d
fi ve
u n
am b
ig u
o u
s o
u tc
o m
es .d
= .0
0 if
n o
n -s
ig n
ifi ca
n t.
N o
t in
G er
sh o
ff
16 Larzelere and Kuhn
T ab
le I.
C o
n ti
n u
ed
B as
is o
f ef
fe ct
si ze
& A
ge ,G
en d
er D
is ci
p li
n e
E ff
ec t
d is
cr ep
an ci
es fr
o m
S tu
d y
(P ar
en t)
a S
am p
le N
ta ct
ic b
O u
tc o
m e
si ze
c (d
) G
er sh
o ff
(2 00
2)
R ea
so n
in g
A m
b it
io u
s, re
al is
ti c
as p
ir at
io n
s, et
c. −.
09 M
ea n
o f
fi ve
as so
ci at
io n
s (o
n ly
o n
e m
ea su
re o
f re
as o
n in
g) P
ri vi
le ge
re m
o va
l A
m b
it io
u s,
re al
is ti
c as
p ir
at io
n s,
et c.
.0 0
M ea
n o
f fi
ve as
so ci
at io
n s
Is o
la ti
o n
A m
b it
io u
s, re
al is
ti c
as p
ir at
io n
s, et
c. .0
0 M
ea n
o f
fi ve
as so
ci at
io n
s
L o
ve w
it h
d ra
w al
A m
b it
io u
s, re
al is
ti c
as p
ir at
io n
s, et
c. .0
0 M
ea n
o f
fi ve
as so
ci at
io n
s
R et
ro sp
ec ti
ve st
u d
ie s
W at
so n
(1 98
9) 0–
5 at
T 1;
17 at
T 2;
B ,G
(M ,
F )
N at
io n
al M
er it
S ch
o la
rs h
ip fi
n al
is ts
& av
er ag
e te
st -t
ak er
s
25 00
P ar
en ta
lr ep
o rt
o f
sp an
k in
g an
d p
o ss
ib ly
ti m
e o
u t
b ef
o re
ag e
6 (c
u st
o m
ar y
P P
)
S co
re o
n N
at io
n al
M er
it S
ch o
la rs
h ip
T es
t an
d re
p o
rt ed
% ra
n k
in cl
as s
−. 11
M ea
n o
f tw
o co
rr el
at io
n s.
O u
tc o
m e
n o
t in
G er
sh o
ff
P ar
en ta
lr ep
o rt
o f
w it
h d
ra w
al o
f p
ri vi
le ge
s an
d as
si gn
in g
ex tr
a d
u ti
es b
ef o
re ag
e 6
S co
re o
n N
at io
n al
M er
it T
es t
an d
% ra
n k
in cl
as s
−. 19
M ea
n o
f tw
o co
rr el
at io
n s
U n
co n
tr o
ll ed
cr o
ss -s
ec ti
o n
al st
u d
ie s
H al
l( 19
94 )
4– 5
B ,G
(M )
P re
sc h
o o
le rs
fr o
m h
ig h
in co
m e
fa m
il ie
s
41 F
re q
u en
cy o
f sp
an k
in g
o r
sl ap
p in
g o
n C
o n
fl ic
t T
ac ti
cs S
ca le
(C T
S )
it em
(c u
st o
m ar
y P
P )
N u
m b
er o
f re
le va
n t
so lu
ti o
n s
in in
te rp
er so
n al
co n
fl ic
t ta
sk
.7 3
C o
rr el
at io
n .O
u tc
o m
e n
o t
in G
er sh
o ff
R ea
so n
in g
(% o
f m
ax im
u m
p o
ss ib
le sc
o re
o n
th re
e C
T S
it em
s)
N u
m b
er o
f re
le va
n t
so lu
ti o
n s
in in
te rp
er so
n al
co n
fl ic
t ta
sk
.7 3
C o
rr el
at io
n
a A
ge in
ye ar
s u
n le
ss o
th er
w is
e in
d ic
at ed
;M =
m ea
n .T
1 =
T im
e 1,
T 2
= T
im e
2; B
= b
o ys
,G =
gi rl
s; (M
) =
m o
th er
s, (F
) =
fa th
er s,
(B ) =
b o
th .
b P
h ys
ic al
p u
n is
h m
en t
is ca
te go
ri ze
d as
ei th
er co
n d
it io
n al
,c u
st o
m ar
y, se
ve re
,o r
p re
d o
m in
an t
u sa
ge .
c A
p o
si ti
ve d
in d
ic at
es a
b en
efi ci
al as
so ci
at io
n ,
i.e .,
th at
gr ea
te r
u se
o f
th e
d is
ci p
li n
ar y
ta ct
ic is
as so
ci at
ed w
it h
p re
fe ra
b le
ch il
d o
u tc
o m
es ,
e. g.
, th
e ta
ct ic
is as
so ci
at ed
w it
h gr
ea te
r p
ro so
ci al
b eh
av io
r o
r le
ss an
ti so
ci al
b eh
av io
r. A
n eg
at iv
e d
in d
ic at
es a
d et
ri m
en ta
la ss
o ci
at io
n b
et w
ee n
th e
ta ct
ic an
d th
e ch
il d
o u
tc o
m e.
d T
h e
m ea
n p
re -p
o st
ga in
fo r
a ch
il d
-d et
er m
in ed
re le
as e
(1 .2
95 )
w as
su b
tr ac
te d
fr o
m th
es e
d s
fo r
an al
ys es
o f
ef fe
ct si
ze s
o f
p h
ys ic
al p
u n
is h
m en
t (e
.g .,
T ab
le II
).
Physical Punishment vs. Alternative Tactics 17
about discipline. Seven of the 26 studies in this meta-analysis were included only in the Larzelere (2000) review. They were probably excluded from the Gershoff (2002) review due to having unusual child outcome variables (three studies of substance abuse, need for power, or realistic/ambitious aspira- tions), falling outside of her search criteria (two stud- ies), being unavailable via interlibrary loan (1), or be- ing part of a study that was already included (1).
Moderating Variables
The following variables were coded to deter- mine whether they accounted for differences in the effect sizes of physical punishment and alternative disciplinary tactics. A study’s design was coded as either (1) a randomized experiment, (2) a non- randomized study that controlled for initial child mis- behavior with statistical controls or within-subject analyses, (3) a time-ordered study (longitudinal, ret- rospective, or sequential) in which the measure of the disciplinary tactic clearly preceded the child out- come measure (without controlling for initial misbe- havior), or (4) a cross-sectional design, in which the referent periods for disciplinary tactics and the child outcome overlapped in time.
Four types of physical punishment were distin- guished. Conditional spanking was defined as phys- ical punishment that was used primarily to back-up milder disciplinary tactics (e.g., reasoning or time- out), used for defiance, or used in a controlled manner. These definitions of conditional spanking emerged because each type demonstrated more ben- eficial outcomes (or less detrimental outcomes) than other types of physical punishment in at least one study (e.g., Larzelere, Schneider, Larson, & Pike, 1996; Ritchie, 1999; Straus & Mouradian, 1998). Al- though the most optimal usage might incorporate all three definitions, no study explicitly incorporated more than one of these definitions in its measure of physical punishment. Customary physical punish- ment was defined as typical parental usage (e.g., us- age or frequency), without emphasizing its severity or predominance. It could have included severe phys- ical punishment, but only to the extent typical of ordi- nary usage by parents. Overly severe physical punish- ment was based on measures that gave extra points for the severity of physical punishment. Examples in- cluded “shaking” (Burton, Maccoby, & Allinsmith, 1961), “severe spankings” (Sears, Maccoby, & Levin, 1957), or spanking when “so angry that you ‘lost it”’
(Straus & Mouradian, 1998). Finally, predominant use of physical punishment included studies inves- tigating predominant disciplinary tactics (e.g., “the primary disciplinary tactic used”) or proportional us- age (e.g., the proportion of disciplinary incidents for which the parents used physical punishment rather than milder disciplinary tactics).
Outcome variables were grouped into four cat- egories, consisting of compliance; antisocial behav- ior (including substance use and abuse); conscience or resistance to temptation; and positive behaviors, competencies, or emotions. Several analyses com- bined antisocial behavior and conscience into a larger category of misbehavior inhibition, to increase statis- tical power for testing other moderating variables.
Same-source bias was coded when an effect size was based solely on information provided by the same person, as opposed to incorporating distinct sources of information. Two age groups distinguished children averaging older or younger than 7 years at the time of the discipline.
Selected Meta-Analytic Details
Most effect sizes were calculated using Johnson’s (1989) DSTAT program, following Gershoff (2002). When a study had multiple relevant statistics, we selected statistics that minimized the methodological problems noted by Baumrind et al. (2002), including the same-source bias and corre- lational statistics. When a study included several statistics that differed on these characteristics, effect sizes were based on the stronger evidence and also distinguished four types of physical punishment (conditional, customary, severe, and predominant usage). In four studies, the best estimate of the effect size controlled statistically for one or more other variables, such as initial child misbehavior. In those cases, the effect size was based on a standardized regression coefficient or similar statistic, following Glass, McGaw, and Smith (1981). In three of those four studies, the effect size could not be estimated in standard deviations units of the outcome vari- able. Therefore, those studies were coded as using a distinct standard deviation unit typical of covariance- corrected coefficients, following Glass et al. (1981). Because the effect sizes from those three studies did not differ significantly from the other 23 studies, the distinction was dropped for the main analyses.
Effect sizes (ds) were corrected for an up- ward bias in small studies, using Hedges’ correction
18 Larzelere and Kuhn
(Lipsey & Wilson, 2001). For the analyses, each ef- fect size was weighted by a function of its sample size and the inverse of its extremity (Lipsey & Wilson, 2001, p. 49; Shadish & Haddock, 1994, p. 268).
The calculations of effect sizes from three sets of studies warrant additional clarification, which is pro- vided in the appendix (the Roberts series of studies, two Larzelere studies, and Ritchie, 1999). The guid- ing principle was to base effect sizes on equivalent analyses of physical punishment and alternative tac- tics. In addition, Larzelere et al. (1996) and Ritchie (1999) yielded different effect sizes for conditional and customary use of physical punishment.
Two effect sizes were included from a study when they were relevant for distinct cells in a partic- ular analysis. For example, several studies had effect sizes for conditional spanking and for another cate- gory of physical punishment. Those studies then con- tributed two different effect sizes for those two types of physical punishment. This increased the sample size from 26 studies to 32 relevant effect sizes in the initial analyses of type of physical punishment.
Following Hedges (1994), the Q statistic was based on ANOVA sums of squares to test hypothe- ses about whether weighted mean effect sizes var- ied significantly by moderating factors. The Q statis- tic is distributed as χ2 under the null hypothesis. Most analyses of moderating variables had missing cells because all combinations of those factors were not represented by at least one study. Consequently, the Q statistic was based on the Type IV sums of squares in the weighted fixed-effects ANOVA. Sta- tistical tests of whether weighted differential means differed from zero used a z statistic, based on Lipsey and Wilson (2001, p. 115).
RESULTS
Effect Sizes of Physical Punishment by Research Design and Physical Punishment Type
The effect sizes of physical punishment on child outcomes varied significantly by type of physical punishment, Q(3) = 9.80, p < .05, by research de- sign, Q(3) = 28.25, p < .001, and by their interac- tion, Q(5) = 17.82, p < .01. As shown in Table II, weighted mean effect sizes appeared detrimental for severe physical punishment (d = −.22) and predom- inant physical punishment (d = −.21), but were near zero for customary and conditional physical punish- ment (ds = .06 and .05, respectively). Mean effect sizes were apparently detrimental in studies using
correlational designs (d = −.22), approached zero in time-ordered and controlled designs (ds = .10 and −.08, respectively), and were apparently beneficial in studies employing randomized designs (d = .80). The interaction effect was due to the following excep- tions to the usual pattern of effect sizes become less detrimental or more beneficial as design quality im- proved: Customary physical punishment produced its most detrimental effect size in statistically controlled studies rather than in cross-sectional studies, whereas predominant usage yielded its most detrimental ef- fect size in time-ordered designs rather than in cross- sectional studies.
Differential Effect Sizes by Research Design and Physical Punishment Type
The next analysis investigated differential effect sizes by research design and type of physical punish- ment. A differential effect size is the difference be- tween the mean effect size for physical punishment and the mean effect size for alternative disciplinary tactics in the same study using the same method- ology. The results showed that neither design nor the Design × Physical Punishment Type interaction was significant, Q(3) = 4.37 and Q(5) = 8.12, respec- tively. Differential effect sizes varied only by the type of physical punishment, Q(3) = 18.26, p < .001.
Table III shows that differential effect sizes fa- vored physical punishment over alternative tactics when physical punishment was defined as conditional (differential d = .29) or customary (differential d = .14). (For brevity, differential d will be shortened to d from here on, which is what it would be called to describe differences between treatment and con- trol conditions.) In contrast, differential effect sizes favored alternative tactics over both overly severe (d = −.07) and predominant physical punishment (d = −.33).
It is instructive at this point to compare the re- sults of the first two analyses. In Table II, the effect sizes associated with physical punishment varied sig- nificantly by research design and by the Design × Physical Punishment Type interaction. In Table III, however, differential effect sizes (d for physical pun- ishment minus d for alternative tactics) did not vary significantly by research design or by its interaction with physical punishment type. These results indi- cate partial success in reducing confounds associ- ated with correlational evidence by using differential
Physical Punishment vs. Alternative Tactics 19
Table II. Weighted Effect Sizes for Physical Punishment by Research Design and Type of Physical Punishment
Research design Type of physical
punishment Cross-sectional Time-ordered Controlled Randomized Weighted mean d
Predominant −.22 (4) −.58 (1) .05 (1) — −.21 (6) Overly severe −.32 (3) .07 (4) — — −.22 (7) Customary .06 (2) .10 (4) −.19 (3) — .06 (9) Conditional −.13 (2) .30 (2) .68 (2) .80 (4) .05 (10) Weighted mean d −.22 (11) .10 (11) −.08 (6) .80 (4) .00 (32) Note. n of studies in parentheses. A positive effect size (d) indicates that higher physical punishment scores are associated with more beneficial or less detrimental child outcomes than are low scores on physical punishment. All mean ds are weighted by Lipsey & Wilson’s equation (2001, p. 49).
effect sizes of alternative tactics from the same study.
To obtain sufficient statistical power to investi- gate hypothesized moderators of the relative effec- tiveness of physical punishment and alternative tac- tics, the following analyses drop research design as a factor because it was not a significant predictor. Like- wise, predominant usage and severe physical punish- ment were combined into one category in subsequent analyses.
Differential Effect Sizes by Outcome and Physical Punishment Type
The next set of analyses investigated whether the differential effect sizes of physical punishment vs. alternatives varied by type of child outcome. The re- sults indicated that outcome type, physical punish- ment type, and their interaction were significantly related to differential effect sizes: physical pun- ishment type, Q(2) = 41.15; outcome type, Q(2) = 21.15; interaction, Q(3) = 16.95, all ps < .001. As ex- pected, conditional spanking showed a more posi- tive differential effect size (d = .29) than customary physical punishment (d = .14), which, in turn, pro-
duced a more positive differential effect size than severe/predominant physical punishment (d = −.12; see Tables IV–VI). Unexpectedly, effect sizes more strongly favored physical punishment for misbehav- ior inhibition (antisocial behavior and conscience; d = .12) than for either compliance (d = .00) or posi- tive behavior and affect (d = .01). The interaction ef- fect reflected the fact that the differential effect sizes for compliance varied by type of physical punishment much more than for other outcomes. Both the largest negative and the largest positive differential effect sizes occurred for compliance. Severe/predominant physical punishment compared less favorably with alternatives for compliance than for any other out- come, whereas conditional spanking compared more favorably with alternative tactics for compliance than for any other outcome. For most outcomes, differen- tial effect sizes were positive for conditional spank- ing, approached zero for customary physical punish- ment, and were negative for severe or predominant usage.
Two of the weighted means listed above de- pended heavily upon results from the largest study, a retrospective survey of substance abuse in 5044 military personnel (Tennant, Detels, & Clark, 1975). Because of its unusually large sample size, this study
Table III. Weighted Differential Effect Sizes (Physical Punishment Minus Alternative Tactics) by Research Design and Type of Physical Punishment
Research design Type of physical
punishment Cross-sectional Time-ordered Controlled Randomized Weighted mean d
Predominant −.37 (4) −.78 (1) .22 (1) — −.33 (6) Overly severe −.11 (3) .05 (4) — — −.07 (7) Customary −.10 (2) .18 (4) −.06 (3) — .14 (9) Conditional .22 (2) .44 (2) .59 (2) .34 (4) .29 (10)
Weighted mean d −.03 (11) .17 (11) .02 (6) .34 (4) .11 (32) Note. n of studies in parentheses. A positive effect size (d) indicates that physical punishment is associated with more beneficial or less detrimental child outcomes than are alternative tactics in the same studies. All mean ds are weighted by Lipsey & Wilson’s equation (2001, p. 49).
20 Larzelere and Kuhn
Table IV. Effect Sizes of Conditionala Physical Punishment Compared to Alternative Disciplinary Tactics
Mean effect size (d)
Alternative Alternative Conditionala Mean difference in disciplinary tactic Child outcome N studies N children disciplinary tactic physical punishment effect sizes
Compliance Reasoning Noncompliance 31−3 152 −.03 .55 .59∗∗∗b Verbal prohibition Immediate
compliance 11 24 −.15 .02 .17
Threats or verbal power assertion
Stop defiance 13 90 .08 .97 .89∗∗∗
Privilege removal or time out
Subsequent compliance
12 38 .02 .01 −.01
Privilege removal Stop defiance 13 90 −.33 .97 1.30∗∗∗ Time out Stop defiance 13 90 .60 .97 .37 Barrier (room time
out) Compliance to
commands & time out
34−6 52 1.04 .84 −.20
Reasoning plus nonphysical punishment
Subsequent compliance
12 38 −.08 .01 .09
Ignoring Stop defiance 13 90 .02 .97 .95∗∗∗ Love withdrawal Immediate
compliance 11 24 .40 .02 −.38
Restraint, physical power assertion
Stop defiance, compliance to commands & TO
23,6 108 .34 .85 .51∗∗
Child release from time out (TO)
Compliance to commands and to time out
26,7 34 .02 .77 .74∗
Mean for compliance 7 220 .26 .68 .43∗∗b
Antisocial behavior Reasoning Aggression 22,8 96 −.22 .35 .56∗∗ Scolding School
aggression 18 58 −.24 .38 .62∗∗
Privilege removal or time out
Aggression 12 38 .26 .26 −.00
Privilege removal School aggression
18 58 .38 .38 .00
Isolation School aggression
18 58 −.18 .38 .56∗
Reasoning plus nonphysical punishment
Aggression 12 38 .63 .26 −.37
Reasoning or nonphysical punishment
Antisocial, impulsivity
19 744 −.39 -.14 .25∗∗∗
Love withdrawal School aggression
18 58 −.24 .38 .62∗∗
Diverting School aggression
18 58 −.47 .38 .85∗∗∗
Mean for antisocial behavior 3 840 −.35 −.07 .28∗∗∗
Physical Punishment vs. Alternative Tactics 21
Table IV. Continued
Mean effect size (d)
Alternative Alternative Conditionala Mean difference in disciplinary tactic Child outcome N studies N children disciplinary tactic physical punishment effect sizes
Conscience Reasoning Conscience 18 86 .22 −.02 −.24 Scolding Conscience 18 86 −.18 −.02 .16 Privilege removal Conscience 18 86 .22 −.02 −.24 Isolation Conscience 18 86 −.30 −.02 .28 Love withdrawal Conscience 18 86 −.04 −.02 .02 Diverting Conscience 18 86 .32 −.02 −.34
Mean for conscience 1 86 .04 −.02 −.06 Grand mean 9 1050 −.23 .06 .29∗∗∗ Note. In the last three columns, effect sizes that are positive indicate that more beneficial outcomes are associated with greater use of (a) alternative disciplinary tactics or (b) conditional physical punishment, or that (c) more beneficial outcomes are associ- ated with conditional physical punishment than with alternative disciplinary tactics. Effect sizes are based on comparable statis- tics, minimizing the methodological problems noted by Baumrind et al. (2002) whenever possible. Means are weighted by sam- ple size and by effect size extremity according to Lipsey and Wilson’s equation (2001, p. 49), with each study contributing one mean of its relevant effect sizes. Some differential effect sizes are not an exact difference of the tabled entries due to rounding. Studies cited: 1Chapman and Zahn-Waxler (1982), 2Larzelere et al. (1996), 3Ritchie (1999), 4Day and Roberts (1983), 5Roberts (1988), 6Roberts and Powers (1990), 7Bean and Roberts (1981), 8Yarrow et al. (1968), 9Straus and Mouradian (1998). aEither (1) nonabusive backup for milder disciplinary tactics in 2- to 6-year-olds, (2) used specifically for defiance in 3- or 4-year-olds, or (3) used in a controlled manner (not out of control due to anger) with 2–14-year-olds.
bSignificant heterogeneity of the effect sizes contributing to this mean, Q statistic, p < .05. ∗p < .05, significantly different from d = .00, z statistic (only performed in the right-hand column). ∗∗p < .01. ∗∗∗p < .001.
was weighted eight times more than the median- sized study in this meta-analysis. Excluding this study, the weighted mean differential effect size dropped to d = .00 for customary physical pun- ishment and to d = .02 for misbehavior inhibition (antisocial/conscience). To avoid an overly favor- able comparison between customary physical punish- ment and nonphysical punishment based solely on this one study, we repeated all analyses excluding the Tennant et al. (1975) study and report the re- sults separately whenever this exclusion changed the findings.
Tables IV–VI distinguish between antisocial behavior and conscience, even though they were treated as misbehavior inhibition in the above anal- yses due to the small number of studies that in- vestigated conscience. Table IV shows that, com- pared to alternative tactics, conditional spanking was associated with greater reductions in noncompli- ance, d = .43, z = 3.08, p < .01, and antisocial behav- ior, d = .28, z = 4.11, p < .001. Conditional spank- ing did not differ significantly from alternative tac- tics in promoting the development of conscience, but this was based on only one cross-sectional study.
Table V shows that, compared to alternatives, cus- tomary physical punishment was associated with greater reductions in antisocial behavior, but this re- sult depended upon the largest study (Tennant et al., 1975), d = .17 with it, z = 6.56, p < .001; d = .03 without it, z = 0.81, ns. Otherwise, customary phys- ical punishment was not significantly different from alternative tactics in its associations with other out- comes. Table VI shows that, compared to alterna- tives, severe/predominant physical punishment was associated with less compliance, d = −.99, z = −4.43, p < .001, conscience, d = −.36, z = −4.48, p < .001, positive behavior and affect, d = −.36, z = −2.28, p < .05, and antisocial behavior, d = .14, z = 2.23, p < .05.
Differential Effect Sizes by Outcome, Physical Punishment Type, and Other Predictors
Additional analyses investigated whether the same-source bias, the child’s age, or the short- vs. long-term timing of the outcomes influenced differ- ential effect sizes and whether they modified the
22 Larzelere and Kuhn
Table V. Effect Sizes of Customarya Physical Punishment Compared to Alternative Disciplinary Tactics
Mean effect size (d)
Alternative Alternative Customarya Mean difference in disciplinary tactic Child outcome N studies N children disciplinary tactic physical punishment effect sizes
Compliance Reasoning Noncompliance 31−3 152 .07 .04 −.03 Verbal prohibition Immediate
compliance 12 24 −.15 .09 .24
Threats or verbal power assertion
Stop noncompli- ance
13 90 .02 .07 .05
Privilege removal or time out
Subsequent compliance
11 38 −.02 −.05 −.03
Privilege removal Stop noncompli- ance
13 90 .24 .07 −.17
Time out Stop noncompli- ance
13 90 .16 .07 −.09
Ignoring Stop noncompli- ance
13 90 .33 .07 −.26
Love withdrawal Immediate compliance
12 24 .40 .09 −.31
Physical power assertion
Stop noncompli- ance
13 90 .20 .07 −.13
Mean for compliance 3 152 .10 .04 −.06 Antisocial behavior
Reasoning Aggression 11 38 .01 .08 .07 Non-contact
punishment Aggression or
substance abuse
21,4 3594 −.07 .23 .31∗∗∗c
Privilege removal Antisocial behavior or alcohol usage
25,6 3285 −.13 −.10 .03
Sent to room Antisocial behavior
15 785 −.18 −.23 −.05
Mean for antisocial behavior 4 6879 −.10 .06 .17∗∗∗bc Conscience
Reasoning Conscience & compliance
17 90 .06 −.04 −.10
Verbal psychological punishment
Conscience, compliance
17 90 −.11 −.04 .07
Love withdrawal Conscience, compliance
17 90 −.06 −.04 .02
Mean for conscience 1 90 −.04 −.04 −.00 Mental health
Privilege removal Neuroticism, esteem
16 2500 .00 −.03 −.03
Positive behavior and affect Reasoning Solutions &
verbal positives in problem solving
18 41 .59 .28 −.31
Physical Punishment vs. Alternative Tactics 23
Table V. Continued
Mean effect size (d)
Alternative Alternative Customarya Mean difference in disciplinary tactic Child outcome N studies N children disciplinary tactic physical punishment effect sizes
Time out, privilege removal, or restitution
Self-esteem 19 157 −.08 −.10 −.02
Privilege removal Prosocial behavior
16 2500 .00 .00 .00
Privilege removal Academic achievement
16 2500 −.19 −.11 .08
Mean for positive behavior and affect 3 2698 −.08 −.05 .03 Grand mean 9 7281 −.08 .06 14∗∗∗bc Note. See Table IV. Studies cited: 1Larzelere et al. (1996), 2Chapman and Zahn-Waxler (1982), 3Ritchie (1999), 4Tennant et al. (1975), 5Larzelere and Smith (2000), 6Watson (1989), 7Lytton (1977), 8Hall (1994), 9Larzelere et al. (1989). aCustomary physical punishment refers to typical usage by parents, e.g., via measures of usage or frequency without emphasizing severity or predominant usage.
bSignificant heterogeneity of the effect sizes contributing to this mean, Q statistic, p < .05. cAfter dropping the largest study (Tennant et al., 1975), the differential effect size is not significantly different from d = .00 and the remaining effect sizes are homogeneous.
∗p < .05, significantly different from d = .00, z statistic (only performed in the right-hand column). ∗∗p < .01. ∗∗∗p < .001.
main conclusions. For each new predictor, the first question was whether it was associated with differ- ential effect sizes at the zero-order level (e.g., cor- relations). If so, the second question was whether its independent contribution was significant af- ter controlling for outcome and type of physical punishment.
Same-Source Bias
Without controlling for other factors, differen- tial effect sizes favored physical punishment over al- ternative tactics more when the same source of infor- mation was used for parent and child variables than otherwise, d = .15 vs. d = .03, Q(1) = 12.42, p < .001. This difference became non-significant, how- ever, after dropping the largest study (Tennant et al., 1975). The next set of analyses controlled for out- come and type of physical punishment in a three-way analysis of variance. Same-source vs. multiple-source data never predicted differential effect sizes signifi- cantly in those analyses, either as a main effect or in statistical interactions with outcome or physical pun- ishment type. This was true regardless of whether the largest study was included. These results suggest that differential effect sizes were successful in minimizing the same-source bias.
Age
Age also predicted differential effect sizes sig- nificantly by itself. Surprisingly, effect sizes favored physical punishment over alternatives for school-age children (d = .20), but not for preschool children, (d = .00, Q(1) = 27.85, p < .001). This age effect dis- appeared when the largest study was omitted, how- ever (Tennant et al., 1975). After controlling for out- come and type of physical punishment, age was a significant predictor in only one interaction. With Tennant et al. (1975) excluded, the Age × Outcome × Physical Punishment Type interaction was signifi- cant, Q(1) = 6.92, p < .01. This interaction was due to differing effects of the Age × Outcome interac- tion by type of physical punishment. For conditional and customary physical punishment, age never pre- dicted differential effect sizes either as a main effect or in an interaction. For severe or predominant phys- ical punishment, the Age × Outcome interaction was significant, Q(1) = 8.34, p < .01. In general, severe or predominant physical punishment was more detri- mental than alternatives for younger than for older children. The major exception to this was that the most detrimental effect of such physical punishment was on self-esteem in older children, based on one study (Coopersmith, 1967).
24 Larzelere and Kuhn
Table VI. Effect Sizes of Severe or Predominanta Physical Punishment Compared to Alternative Disciplinary Tactics
Mean effect size (d)
Alternative Alternative Severe/Predominanta Mean difference in disciplinary tactic Child outcome N studies N children disciplinary tactic physical punishment effect sizes
Compliance Reasoning Noncompliance 11 90 .44 −.55 −.99∗∗∗
Antisocial behavior Reasoning Antisocial or
need for power
22,3 116 −.25 .30 .55∗∗
Privilege removal or time out
Antisocial behavior
12 38 .20 .05 −.15
Privilege removal Aggression or need for power
23,4 238 −.08 .23 .31∗
Reasoning plus nonphysical punishment
Antisocial behavior
12 38 .10 .05 −.05
Reasoning or nonphysical punishment
Antisocial, impulsivity
15 744 −.39 −.28 .11
Love withdrawal Aggression or need for power
23,4 238 .08 .23 .15
Mean for antisocial behavior 4 1020 −.29 −.15 .14∗ Conscience
Reasoning Conscience or resistance to temptation
26,7 402 .27 −.34 −.61∗∗∗b
Scolding Conscience or resistance to temptation
17 76 .08 −.09 −.17
Privilege removal Conscience or resistance to temptation
36−8 506 −.14 −.33 −.19∗b
Isolation Conscience or resistance to temptation
36−8 506 .01 −.33 −.33∗∗∗
Love withdrawal Conscience or resistance to temptation
36−8 506 .22 −.33 −.54∗∗∗
Love withdrawal and reasoning
Internal moral motivation
19 120 .00 −.16 −.16
Mean for conscience 4 626 .07 −.30 −.36∗∗∗ Positive behavior and affect
Reasoning Prosocial behavior, self-esteem, aspirations
310−12 162 .31 −.20 −.51∗∗b
Verbal prohibition Prosocial behavior
112 16 −.70 −.58 .12
Physical Punishment vs. Alternative Tactics 25
Table VI. Continued
Mean effect size (d)
Alternative Alternative Severe/Predominanta Mean difference in disciplinary tactic Child outcome N studies N children disciplinary tactic physical punishment effect sizes
Privilege removal Aspirations (realistic, yet ambitious)
111 83 .00 .00 .00
Isolation Aspirations (realistic, yet ambitious)
111 83 .00 .00 .00
Ignoring Prosocial behavior
112 16 −.34 −.58 −.24
Restraint Prosocial behavior
112 16 .26 −.58 −.84
Restraint, denial, or isolation
Self-esteem 110 63 .86 −.42 −1.29∗∗∗
Love withdrawal Self-esteem, aspirations
210,11 146 −.23 −.18 .05
Mean for positive behavior and affect 3 162 .15 −.21 −.36∗b Grand mean 12 1898 −.10 −.22 −.12∗b Note. See Table IV. Studies cited: 1Minton et al. (1971), 2Larzelere et al. (1998), 3McClelland and Pilon (1983), 4Sears (1961), 5Straus and Mouradian (1998), 6Sears et al. (1957), 7Burton et al. (1961), 8Grinder (1962), 9Aronfreed (1961), 10Coopersmith (1967), 11Crowne et al. (1969), 12Zahn-Waxler et al. (1979). aSevere physical punishment indicates that the measure included extra points for its severity in addition to its frequency or usage; pre- dominant use indicates either that physical punishment was the major tactic used or that it was assessed with a proportional usage measure.
bSignificant heterogeneity of the effect sizes contributing to this mean, Q statistic, p < .05. ∗p < .05, significantly different from d = .00, z statistic (only performed within the right-hand column). ∗∗p < .01. ∗∗∗p < .001.
Short-Term vs. Long-Term Outcomes
Whether the outcomes were assessed short- term or long-term was completely confounded with research design. Randomized, cross-sectional, and sequential within-subject designs yielded outcomes ranging from immediately to the next day, whereas other designs investigated outcomes at least two months after the disciplinary tactics were used. By itself, the timing of the outcome was significantly re- lated to differential effect sizes only when the Ten- nant et al. (1975) study was included in the analyses, Q(1) = 7.44, p < .01. In that case, differential effect sizes favored physical punishment over alternatives in long-term outcomes (differential d = .12), but not in short-term outcomes (d = .00). After controlling for outcome and type of physical punishment, differ- ential effect sizes continued to favor physical pun- ishment over alternatives more for long-term out- comes, but only when Tennant et al. was included, Q(1) = 9.71, p < .05.
Comparisons with Specific Alternative Tactics
The analyses to this point have compared phys- ical punishment to all alternative tactics together. This grouping could obscure distinctions among al- ternative tactics in how favorably they compare with physical punishment. Accordingly, Tables IV– VI summarize the weighted mean effect sizes for specific alternative tactics, compared to conditional spanking (Table IV), to customary physical punish- ment (Table V), and to severe or predominant phys- ical punishment (Table VI).
Overall, conditional spanking was associated with better child outcomes than were alternative disciplinary tactics, d = .29, p < .001 (Table IV), but this applied only to reductions in noncompliance and antisocial behavior. Conditional spanking did not differ from any alternative tactic in its association with conscience, according to the only relevant study (Yarrow et al., 1968; overall d = −.06, ns). Specific differential effect sizes favored conditional spanking
26 Larzelere and Kuhn
over 10 of 13 alternative tactics in associations with noncompliance, antisocial behavior, or both. The dif- ferential effect sizes ranged from 1.30 to −.38, de- pending upon the specific alternative tactic and the outcome. In no case was an alternative tactic associ- ated with significantly less noncompliance or antiso- cial behavior than conditional spanking.
Combining data for all outcomes, the weighted mean effect sizes for specific tactics favored only 2 of 13 alternative tactics over conditional spanking, al- beit non-significantly so. The barrier-enforced back- up for time-out had a differential mean d of −.20, z = −.74, ns. The combination of nonphysical pun- ishment and reasoning4 had a differential effect size of d = −.02, ns (Larzelere et al., 1996). Verbal prohi- bition was the only other disciplinary tactic that did not differ significantly from conditional spanking in its associations with child outcomes. Other mean dif- ferential ds ranged from .17 (love withdrawal,5 ver- bal prohibition) to .95 (ignoring), after combining weighted means across all outcomes in Table IV.
Table V shows that customary physical punish- ment averaged being more effective than alterna- tive disciplinary tactics overall, d = .14, z = 5.84, p < .001. The mean differential effect size would have been d = .00, except for the large retrospective study of substance abuse (Tennant et al., 1975), which had a moderately large differential effect size favoring physical over nonphysical punishment, d = .31, p < .001. All the significant differential effect sizes for customary physical punishment depended upon the large Tennant et al. (1975) study for its significance. Otherwise, no tactic differed significantly from cus- tomary physical punishment in its effect size with any outcome, in either direction.
Alternative tactics tended to be associated with better outcomes than severe or predominant physi- cal punishment, although by a surprisingly small de- gree, d = −.12, z = −2.49, p < .05. Specific tactics that produced significantly better effect sizes than severe/predominant physical punishment for at least one outcome included reasoning, time-out (termed “isolation”6 in some older studies), and love with-
4Reasoning is defined as verbal attempts to persuade a child to behave appropriately, e.g. using description of consequences or another type of explanation.
5Love withdrawal is defined as temporarily withholding expres- sions of love and nurturance from the child, used as a discipline category in 11 of the studies, from 1957 to 1983.
6Isolation is a term used for a disciplinary tactic in older stud- ies, which seems to approximate time-out better than other disci- plinary categories from that era.
drawal. Privilege removal had a significantly better effect size than severe/predominant physical punish- ment for enhancing conscience, but a significantly worse effect size for reducing antisocial behavior. The other six disciplinary tactics never differed sig- nificantly from severe/predominant physical punish- ment in their associations with any outcome, al- though the direction of their differential effect sizes favored five alternatives over severe/predominant physical punishment.
DISCUSSION
The major methodological goal of this meta- analysis was to reduce several confounds in this pre- dominantly correlational literature by analyzing dif- ferences in effect sizes between physical punishment and alternative disciplinary tactics. The fact that dif- ferential effect sizes did not differ by factors such as research design and the same-source bias suggests that this goal was achieved, at least in part. We then evaluated previous conclusions about physical pun- ishment by investigating whether detrimental child outcomes associated with physical punishment are distinctive of physical punishment itself or whether similar outcomes are also found for alternative disci- plinary tactics, when investigated with the same re- search methods.
Results by Type of Physical Punishment
Whether physical punishment compared favor- ably or unfavorably with other tactics depended on the type of physical punishment. Conditional spank- ing was more strongly associated with reductions in noncompliance or antisocial behavior than 10 of 13 alternative disciplinary tactics. The mean effect size differences favored conditional spanking over alternatives by amounts between Cohen’s (1988) small and medium effect sizes, sizable differences for comparisons between alternative treatments such as these (d = .43 for noncompliance and d = .28 for an- tisocial behavior). The effect sizes of customary phys- ical punishment were neither worse nor better than any alternative tactic, with the exception that Ten- nant et al. (1975) found physical punishment to be as- sociated with less substance abuse than non-contact punishment. The differential effect sizes favored al- ternative tactics only in comparisons with overly se- vere or predominant use of physical punishment.
Physical Punishment vs. Alternative Tactics 27
Conditional spanking compared favorably with alternatives when defined by one of three crite- ria. Conditional spanking compared most favorably with alternative tactics when it was defined as a response to defiance in 3- and 4-year-olds (mean d = .54: Ritchie, 1999; Yarrow et al., 1968) or as a back-up for time-out in clinically oppositional 2–6- year-olds (mean d = .32 from Roberts’s four stud- ies). Conditional spanking also compared favorably with alternatives when it was defined as controlled usage, not out of control due to anger (mean d = .25: Straus & Mouradian, 1998; see also Turner & Muller, 2004). On a fourth criterion, combining spanking with reasoning, the outcomes of conditional spank- ing were no better than alternative tactics (mean d = .02: Chapman & Zahn-Waxler, 1982; Larzelere et al., 1996). It should also be noted that the studies of con- ditional spanking involved children between 2 and 6 years old, except for Straus and Mouradian (1998), whose sample was 2–14 years of age.
Physical punishment had effect sizes more detri- mental than alternatives only when it was used severely or as the predominant disciplinary tactic. This finding supports the consensus statement from the 1996 scientific consensus conference on physical punishment that “Spanking a child should not be the primary or only response to a misbehavior used by a care giver” (Friedman & Schonberg, 1996a, p. 853). It also extends that conference’s eighth consensus statement by showing increased risk of dysfunction from severe physical punishment, regardless of the age of the child.
Next, we consider the implications of these re- sults for specific previous conclusions about physical punishment, followed by implications for other disci- plinary tactics.
Variations by Outcomes, Duration, and Alternative Tactics
The results of this meta-analysis contradict some previous conclusions about physical punishment and support other conclusions. First, the findings contra- dict previous conclusions that the only exception to the detrimental effects of physical punishment is for immediate child compliance (Gershoff, 2002; Straus, 2001). Instead, the effect sizes of conditional spank- ing compared favorably with alternative tactics for all disruptive behavior problems, including antisocial behavior and defiance. Second, physical punishment competed just as well with alternative tactics for long-
term outcomes as for short-term outcomes. In fact, the results favored physical punishment over alter- natives more for long-term outcomes than for short- term outcomes, but only when the largest retrospec- tive study (Tennant et al., 1975) was included in the analyses.
Third, all types of physical punishment were associated with lower rates of antisocial behavior than were alternative disciplinary tactics. Condi- tional spanking produced effect sizes more favor- able than alternative tactics for subsequent school aggression in 4-year-olds (Yarrow et al., 1968) and for concurrent antisocial behavior in 2–14-year-olds (Straus & Mouradian, 1998). Customary physical punishment was associated with lower substance abuse than were other tactics (Tennant et al., 1975). Even overly severe or predominant physical punish- ment predicted less antisocial aggression than did alternative tactics, based on two longitudinal stud- ies (McClelland & Pilon, 1983; Sears, 1961) and one cross-sectional study (Straus & Mouradian, 1998). Four other studies found that physical punishment and alternative tactics did not differ in their associ- ations with antisocial behavior. Thus, if physical pun- ishment increases aggression and antisocial behavior, it does so to the same degree or less than the disci- plinary tactics to which it has been directly compared.
Fourth, this meta-analysis failed to detect neg- ative side effects unique to physical punishment (Benjet & Kazdin, 2003). This finding is consistent with three reviews that concluded that the nega- tive side effects of punishment are minor and can easily be avoided by making the punishment con- tingencies clear and by reinforcing appropriate be- havior (Matson & Taras, 1989; Newsom, Favell, & Rincover, 1983; Walters & Grusec, 1977). Indeed, two reviews (Matson & Taras, 1989; Newsom et al., 1983) concluded that positive side effects of punish- ment were more common than negative side effects. One review (Walters & Grusec, 1977) considered physical aggression to be an established negative side effect of physical punishment. On that point, this meta-analysis found no evidence that physical punishment was more strongly associated with phys- ical aggression than other disciplinary tactics.
Two previous conclusions about physical pun- ishment were partially supported. The first was that physical punishment fails to teach positive alter- native behaviors. No form of physical punishment was more strongly associated with the development of conscience or of positive behaviors, emotions, or competencies than were alternative tactics. Yet,
28 Larzelere and Kuhn
compared with other disciplinary tactics, physical punishment predicted lower positive child outcomes only when it was used severely or predominantly, similar to the results for other outcomes.
Second, the analyses partially supported the conclusion that nonphysical punishments are just as effective as physical punishment. The strongest com- petitors of conditional spanking for reducing be- havior problems included the barrier-enforcement method (“room time-out”: Day & Roberts, 1983; Roberts, 1988; Roberts & Powers, 1990), a com- bination of nonphysical punishment and reasoning (Larzelere et al., 1996), and verbal prohibition (only in Chapman & Zahn-Waxler, 1982). In addition, time-out (termed “isolation” in older studies) was equally effective at inhibiting misbehavior in con- trasts with conditional and customary physical pun- ishment combined.
The outcomes of nonphysical punishment were not always equivalent to those of physical punish- ment, however. The meta-analytic results favored conditional spanking over nonphysical punishments in general for reducing defiance and antisocial be- havior, mean d = .34, p < .05. In addition, custom- ary physical punishment was associated with less substance abuse than was non-contact punishment (Tennant et al., 1975).
To summarize this section, this meta-analysis only partially supported prevailing conclusions about physical punishment. Most of the previous evidence against physical punishment does not appear to be unique to physical punishment. Equivalent analyses produce similar evidence against a range of alterna- tive disciplinary tactics as well. This pattern is what would be expected if systematic biases, such as the intervention selection bias, the same-source bias, and the lumping bias, account for most of the evidence against physical punishment. Two previous conclu- sions were supported in this meta-analysis. First, physical punishment, like other forms of punishment, does not enhance positive development, but only in- hibits inappropriate behavior, such as defiance and antisocial behavior. Second, most types of nonphys- ical punishment had similar associations with out- comes as did physical punishment, although they had better outcomes only in comparisons with overly se- vere or predominant physical punishment.
If differential effect sizes provide only partial support for prevailing conclusions about physical punishment, what are the implications for research on parental discipline in general? The next section considers these broader implications.
Implications for Parental Discipline Research
The most controversial aspect of parental disci- pline is whether any kind of punishment ever has a place in optimal parental discipline. If so, what dis- tinguishes effective from counterproductive uses of punishment? Two distinct empirical traditions have answered these questions differently. Developmental psychologists tend to view all forms of punishment negatively, especially compared with disciplinary reasoning (Bee, 1998; Berger & Thompson, 1995; Bornstein & Lamb, 1988; Etaugh & Rathus, 1995; Grolnick, Deci, & Ryan, 1997; Grusec, 1997; Holden, 1997; Kochanska, Padavich, & Koenig, 1996). In con- trast, behavioral parent trainers typically teach non- physical punishment as a core skill and discourage verbal discipline except to clarify instructions and contingencies (Aronfreed, 1968; Axelrod & Apsche, 1983; National Institutes of Health, 1991; Patterson, 1982; Walters & Grusec, 1977). Most developmen- tal psychologists might predict reasoning to have consistently better outcomes than physical punish- ment, whereas behavioral parent trainers would ex- pect nonphysical punishment to compare favorably with both reasoning and physical punishment. In this meta-analysis, reasoning and nonphysical pun- ishment not only failed to have better outcomes than conditional and customary physical punishment, but neither one was consistently better than the other in competing with physical punishment across all out- comes.
The fact that neither perspective was entirely confirmed by these results suggests the need to im- prove both perspectives. The two viewpoints com- plement each other very well in most respects (Larzelere, 2001; Larzelere et al., 1996), and their complementary strengths suggest some constructive directions for future conceptualization and research. The developmental psychology perspective tends to be stronger on ecological validity and has several fo- cused theories with the potential to integrate the two perspectives constructively. Behavioral parent train- ing is stronger on internal validity and on specifying effective implementation of many parental discipline skills.
The respective strengths of both perspectives have corresponding weaknesses. Because most re- search in development psychology uses correlational methodology, it tends to be weak in internal validity. Such research must rule out plausible alternative ex- planations to determine whether their apparent ev- idence against punishment can be accounted for by
Physical Punishment vs. Alternative Tactics 29
child effects, genetic effects, and other non-parental effects. The main deficiency in behavioral parent training is that it does not explain how parent–child interaction can best move beyond clarifying and en- forcing contingencies. The correlational evidence in developmental psychology shows that predominantly verbal discipline and well-behaved children tend to go together, even though such correlations may be inadequate for confirming the bidirectional causal effects that lead to that happy conclusion. Behav- ioral parent training may show how parents can re- gain reasonably cooperative interactions with oppo- sitional children, but the field has not been as strong at identifying subsequent links in the causal chain to- ward optimal verbal parent–child interaction, espe- cially of the type desired as children grow toward adulthood.
We suggest that the strengths of both perspec- tives could be integrated to advance our understand- ing of parental discipline. This integrated model must be able to distinguish effective from counterproduc- tive disciplinary tactics on the basis of a number of factors in addition to the type of disciplinary tactic. These additional factors might include the parent– child context, characteristics of the child and the sit- uation, and the discriminations made by parents as they make ongoing discipline choices.
Integrative Theories
Several focused theories in developmental psy- chology suggest complementary roles for disciplinary reasoning and punishment, but their integrative potential has not been fully exploited (Grusec, 1997). According to Baumrind’s (1973, 1991) parent- ing styles, authoritative parenting produces optimal child outcomes with a combination of firm control, nurturance, and good communication. Hoffman’s (1977) information processing theory emphasizes dis- ciplinary reasoning combined with an intermediate level of disciplinary firmness. Bell and Harper (1977) provided evidence that parental discipline operates like a control system, in which parents use correc- tive discipline to get children to return to an ac- ceptable range of behavior. Larzelere’s (2001) con- ditional sequence model applied this control system model to a sequential choice of disciplinary tactics, wherein milder disciplinary tactics are preferred, but are backed up with increasingly forceful tactics when child noncompliance persists beyond what is accept- able to the parent.
Such integrative views of reasoning and pun- ishment can account for the otherwise puzzling re- sults in this meta-analysis and in an article by Grusec and Goodnow (1994). The significance of the Grusec and Goodnow article is that they documented the in- consistent empirical evidence for the presumed su- periority of disciplinary reasoning over punishment. Reasoning is associated with better outcomes than punishment in middle-class families, but rarely in working-class families, preschoolers, boys, or tem- peramentally difficult children. The latter samples all have a higher proportion of children who challenge disciplinary limits, thus eliciting more forceful en- forcement tactics. This pattern of results is consis- tent with the view that disruptive children require more frequent use of punishment to support the ul- timate goal of maintaining appropriate cooperation with verbal discipline.
Three findings of this meta-analysis are con- sistent with these integrative theories. First, phys- ical punishment was associated with better out- comes than most alternative tactics only when it was limited to controlled spanking for defiant re- sponses to milder disciplinary tactics. This imple- mentation of spanking enforces cooperation with milder disciplinary tactics, thereby supporting the goal of appropriate cooperation using milder verbal tactics. Second, physical punishment was associated with better outcomes than alternatives only for re- ductions in noncompliance and antisocial behavior. Compared to alternatives, no type of physical punish- ment was associated with positive outcomes, such as conscience development and positive behaviors and feelings.
Third, the relative effect sizes of reasoning and nonphysical punishment suggest they may play com- plementary roles, consistent with these integrative theories. When reasoning and nonphysical punish- ment were both compared with physical punishment (Tables IV–VI), reasoning was more effective than nonphysical punishment for enhancing positive child characteristics, but nonphysical punishment was bet- ter for inhibiting misbehavior.
Effective Tactic Implementation
Along with integrating theories, future research is needed to investigate how parents can use spe- cific disciplinary skills more effectively. Helping par- ents to skillfully encourage appropriate behavior, prevent discipline problems, and respond to them
30 Larzelere and Kuhn
with effective verbal correction should reduce the need for punishment of any kind. Further, more skillful use of nonphysical punishment should re- duce the need for physical punishment. The most impressive reductions in physical punishment have occurred after parents were trained in a range of dis- ciplinary skills, including effective use of a last-resort tactic, whether conditional spanking (Eyberg, 1993; Roberts, 1984) or alternative ways to enforce com- pliance with time out (McNeil, Clemens-Mowrer, Gurwitch, & Funderburk, 1994; Webster-Stratton, 1990). The frequency of physical punishment de- creased far more from these interventions than from laws or recommendations against spanking (Ispa & Halgunseth, 2004; Statistics Sweden, 1996). There- fore, training parents in a range of disciplinary skills may be more effective for both child and parenting outcomes than merely prohibiting traditional last- resort tactics.
Variations in Effectiveness by Situation
Finally, understanding parental discipline would be enhanced by understanding the discriminations by which parents choose one disciplinary tactic over another. Effective parents do not rely on the same disciplinary tactics all the time, but match them to the situation. In Ritchie’s (1999) study, mothers of 3-year-olds were most likely to select spanking and time out for defiance, privilege removal for simple refusals or passive noncompliance, and ignoring for passive noncompliance or whining (Larzelere, 2002). In each case, the preferred tactic was optimally effec- tive for terminating that particular form of noncom- pliance. These findings suggest that mothers choose disciplinary tactics based on finer discriminations than those represented in theories about parental dis- cipline. Perhaps the immediate effectiveness of some of these tactics is outweighed by detrimental long- term effects. This meta-analysis, however, indicates that distinctive long-term effects of physical punish- ment are similar to its short-term effects.
In any case, parents generally modify their dis- ciplinary tactics based on the perceived effects, but using a much quicker feedback loop than reflected in most research studies. Ritchie (1999) showed that mothers modified their disciplinary tactics when noncompliance persisted within a single disciplinary episode, changing somewhat from verbal tactics to disciplinary punishment. Roberts and Powers (1990) demonstrated the effectiveness of modifying a last-
resort tactic with clinically defiant children. If the children did not comply with time out after six rep- etitions of the initial back-up tactic, they switched to an alternative back-up tactic, changing from spank- ing to the barrier-enforcement, or from any other back-up to spanking. In each case, compliance to time out was achieved with the second back-up tac- tic. The importance for parents to adjust their dis- ciplinary response to the situation was emphasized in the following quote from Grusec and Goodnow (1994):
Hoffman (1970) [observed] that different situations seemed to ‘pull’ a particular type of discipline from the parent and that this variation of discipline tech- nique by the situation was particularly the case among mothers of children who had a strong moral orientation. The minimal implication is that one may need to look at the nature of the misdeed, and at the connection between misdeed and disciplinary tech- nique, as a part of the answer to how and when dif- ferential effectiveness occurs. The large-scale impli- cation is that explanations should be directed toward accounting for why flexibility is effective, rather than being directed only toward the differential effective- ness of the methods themselves (p. 7).
In summary, research on parental discipline gen- erally needs to account for the inconsistency of the evidence both for the superiority of recommended disciplinary tactics and for the inferiority of disfa- vored disciplinary tactics. Consistent with Grusec and Goodnow’s points (1994), future research needs to account for child effects, genetic effects, and on- going bidirectional influences between children and parents; it needs to make finer discriminations rather than lumping broad categories of disciplinary tac- tics together; and it needs to account for how par- ents vary their disciplinary tactics according to the situation, the type of misbehavior, the child’s ini- tial response, and other factors. Otherwise, simplistic methods will continue to yield simplistic conclusions that fail to do justice to the complexity of parental discipline.
Limitations
The quality of any meta-analysis depends, of course, on the quality and quantity of the qualify- ing studies. Indeed, eighteen of the 26 studies in this meta-analysis relied on weak methodologies, providing only zero-order correlations from cross- sectional, retrospective, or longitudinal designs. The other eight studies included four randomized studies,
Physical Punishment vs. Alternative Tactics 31
two longitudinal studies with statistical controls, and two that used within-subject analyses. The random- ized studies had the smallest sample sizes, with only eight or nine children per disciplinary tactic. Weight- ing studies by sample size tended to minimize the in- fluence of these randomized studies relative to the larger, but poorly controlled studies. Moreover, even though this meta-analysis reduced some biases by an- alyzing differences between the effect sizes of physi- cal punishment and alternative disciplinary tactics, it could not eliminate systematic biases entirely.
Altogether, only 26 studies were found that in- vestigated physical punishment and an alternative tactic in children under the age of 13. Consequently, the mean effect sizes in some cases were based on only one or two studies. The number of studies was insufficient to analyze all relevant predictors simulta- neously, necessitating the step-by-step consideration of hypothesized moderators of differences in effect sizes.
Potential Concerns
Given the importance of this topic for children’s welfare, it is important to consider the plausibility of alternative explanations for our meta-analytic con- clusions. For a predominantly correlational research literature, causal evidence can be supported only to the extent that other plausible explanations have been ruled out (Larzelere et al., 2004; Shadish et al., 2002, chap. 14). The next section considers several al- ternative interpretations.
Escalation Toward Abuse
One important concern is the possibility of es- calation from mild spanking to severe physical pun- ishment. Virtually all professionals oppose overly severe physical punishment, a view consistent with these meta-analytic results. The point of this con- cern, however, is the inadequacy of evaluating con- ditional spanking only when it remains within its de- fined parameters (e.g., used in a controlled manner). Parents may intend to spank in a controlled manner, yet end up inflicting more pain than intended. This type of escalation is not represented in the effect size of conditional spanking once it crosses that line. The escalation issue is mitigated somewhat by the fact that customary physical punishment never compared unfavorably with any alternative tactic, because
customary physical punishment includes escalations to the degree that they occur in typical parental discipline.
Nonetheless, better research is needed on es- calation processes within parent–child disciplinary interactions. Most accused parents portray physi- cally abusive incidents as emanating from a disci- pline incident (Kadushin & Martin, 1981). We know little about escalation processes within discipline incidents, however. Gershoff (2002) found a consis- tently positive association between physical punish- ment and physical abuse, but that necessarily fol- lows from defining all physical abuse as instances of physical punishment (100% of abusers then used physical punishment compared to a lower percent- age of non-abusers). Aside from that correlation, the linkage is based on the domino fallacy (Baumrind, 1983; Damer, 1980), which holds that any step in an undesirable direction (e.g., spanking or buying on credit) is always undesirable because it increases the possibility of its undesirable extreme (abuse or bankruptcy). There are clearly some types of phys- ical punishment that increase the risk of abuse, but it is not clear that all forms of physical punishment increase that risk.
There is evidence that the likelihood of dis- ciplinary escalation is reduced when milder dis- ciplinary tactics become more effective. Several studies have shown that milder disciplinary tac- tics become more effective after they are consis- tently enforced with nonphysical punishment and, if necessary, non-abusive spanking (Larzelere, Sather, Schneider, Larson, & Pike, 1998; Roberts & Powers, 1990). Consequently, one way to reduce the risk of escalation toward abuse is to help parents use milder disciplinary tactics more effectively.
Research has yet to show that a ban on spanking reduces the subsequent rate of physical abuse by par- ents (Larzelere & Johnson, 1999). Along with other possibilities, it may be that a spanking ban eliminates the kind of age-appropriate nonabusive spanking that helps parents enforce milder disciplinary tactics, thereby increasing the risk of escalating disciplinary interactions. More objective evaluations of spanking bans are needed before they become models for uni- versal dissemination.
Controlled Longitudinal Studies
A second potential concern is that the studies with the strongest causal evidence against physical
32 Larzelere and Kuhn
punishment were not represented in this meta- analysis, including five longitudinal studies that con- trolled statistically for the initial level of the outcome. Straus (2001) has argued that these studies compel all professionals to oppose all spanking as a disci- plinary option for parents. Unfortunately, these stud- ies did not qualify for this meta-analysis because they did not investigate an alternative disciplinary tactic. Larzelere and Smith (2000), however, took advan- tage of the fact that the longitudinal cohort analyzed by Straus, Sugarman, and Giles-Sims (1997) also included parallel questions about four alternative tactics. When analyzed in the same manner that they analyzed spanking, the four alternative disciplinary tactics also predicted higher subsequent antisocial behavior, significantly so for grounding, marginally for privilege removal and allowance removal, and non-significantly for sending children to their room. In addition, the effects for spanking and all four alter- native tactics became non-significant when the mea- sure of initial antisocial behavior was improved from the trichotomous measure used by Straus et al. (1997) to a continuous measure of externalizing behavior problems.
These findings suggest that child effects (i.e., the intervention selection bias) were only partially con- trolled in Straus et al. (1997), because their covari- ate distinguished only among zero, low, and high levels of initial antisocial behavior. Epidemiologists have shown that residual confounding remains when a covariate is converted into a dichotomous or tri- chotomous variable (Rothman & Greenland, 1998). Similar types of inadequate statistical controls led ini- tially to an overly negative evaluation of Head Start (Campbell & Boruch, 1975; Campbell & Erlebacher, 1970).
The adequacy of the other four controlled longi- tudinal studies cited by Straus (2001) falls well short of Straus et al. (1997). Two investigated slapping (Brezina, 1999) or spanking teenagers (Simons, Lin, & Gordon, 1998), one failed to control for the ini- tial level of misbehavior (Straus & Paschall, 1998), and the other concluded prominently, “For most chil- dren, claims that spanking teaches aggression seem unfounded” (Gunnoe & Mariner, 1997, p. 768). The results of these controlled longitudinal studies are mixed at best, with the strongest evidence against customary spanking applying almost as strongly to other disciplinary tactics. Moreover, the strongest evidence disappears with improved controls for ini- tial level of misbehavior. Nonetheless, more studies are needed that control for initial child misbehav-
ior,7 preferably ones that compare the outcomes of nonabusive spanking directly with alternative tactics.
Equivalent Effects
A third objection might be based on the fact that two forms of nonphysical punishment yielded effect sizes equivalent to conditional spanking. The availability of an equally effective nonphysi- cal punishment has been one rationale for dispens- ing with physical punishment altogether (Graziano, Hamblen, & Plante, 1996; Straus, 2001). Equivalence of effects makes a poor rationale for a spanking ban for several reasons. Disciplinary tactics with equiv- alent effectiveness overall may each show superior effectiveness for some children in some situations. Indeed, the barrier method, the most effective dis- ciplinary tactic in this meta-analysis, was ineffective with some children, and a child-determined release from time-out, a relatively ineffective disciplinary tactic, was effective for some clinically oppositional children (Roberts & Powers, 1990). When one disci- plinary tactic is not working, parents would benefit from having a range of effective alternatives to turn to, as shown by Roberts and Powers (1990). More- over, effect equivalence would not be considered suf- ficient to ban a prescription medication, unless the differences in negative side effects clearly favored other equally effective medications in almost all applications.
Conditional Spanking vs. Non-Optimal Alternatives
A final anticipated objection might be that the current meta-analysis is biased in favor of conditional spanking, because it was not compared with optimal forms of alternative disciplinary tactics. This criti- cism would be appropriate, in part, for two defini- tions of conditional spanking (controlled usage or combined with reasoning), but not for the others (defiance, enforcing time-out). Controlled spanking was compared with alternative disciplinary tactics, regardless of whether they were used in a controlled manner (Straus & Mouradian, 1998). But even out- of-control physical punishment compared favorably
7As this article was going to press, Grogan-Kaylor (2004) pub- lished a study of customary spanking that used an econometric longitudinal analysis to control for omitted as well as incorpo- rated variables.
Physical Punishment vs. Alternative Tactics 33
with recommended alternatives in that study, al- though not to the same degree as controlled spank- ing. As for spanking in combination with reasoning, it was compared with a range of tactics, only one of which was a parallel combination of nonphysical pun- ishment and reasoning (Larzelere et al., 1996).
In the other two definitions of conditional spanking, alternative tactics were compared with spanking under identical conditions, either respond- ing to defiance (Ritchie, 1999; Yarrow et al., 1968) or to non-compliance with time-out (four studies by Roberts and his colleagues). Note that the effect sizes favored conditional spanking the most under the lat- ter two definitions, where the operational definition did not entail a biased comparison with alternative tactics.
CONCLUSION
The results of this meta-analysis indicate that the detrimental child outcomes previously associated with physical punishment are not unique to physical punishment itself, except when it is used severely or predominantly. In fact, conditional spanking was as- sociated with significantly less defiance or antisocial behavior than 10 of 13 alternative disciplinary tac- tics. The usual ways that parents use physical pun- ishment (“customary”) were never associated with worse outcomes than any alternative tactic. A pos- sible explanation for these results is that systematic methodological biases account for the correlational evidence linking physical punishment to detrimental child outcomes. When the same methods and logic are used for alternative disciplinary tactics, child out- comes appear equally detrimental.
We are not the first to point out the weak and uneven empirical support underlying prevailing con- clusions about parental discipline (Grusec & Good- now, 1994; Harris, 1998). For recommended disci- plinary practices to be based on scientific evidence, researchers must account for the inconsistency of em- pirical support for prevailing conclusions. Integrative theories can account for the inconsistent data, while reconciling disparate views about disciplinary tactics from two distinct empirical literatures.
Before social scientists conclude that the 94% of American mothers who spank their 3- and 4- year-olds are invariably doing harm to their chil- dren (Straus & Stewart, 1999), they should have sup- portive evidence that is distinctive of nonabusive spanking. Psychologists owe it to parents and their
children to base recommendations for sweeping so- cietal changes on the best scientific evidence possi- ble. Without that safeguard, children are at risk for becoming the victims of well intentioned, but prema- ture policy changes.
APPENDIX: SELECTED EFFECT SIZE DETAILS
The appendix first compares the current study’s effect sizes with those estimated by Gershoff (2002). It then provides additional specifics about our effect size estimates from three sets of studies.
Comparisons of Effect Sizes with Gershoff (2002)
The overall mean weighted effect size for physi- cal punishment in the current meta-analysis was d = .00, compared to Gershoff’s (2002) d = −.35, which indicated somewhat detrimental associations with child outcomes. The seven studies herein that were not included in Gershoff had a weighted mean of d = .15, indicating slightly beneficial associations of phys- ical punishment with child outcomes. The overall beneficial association was produced primarily by the largest study, a retrospective analysis of childhood predictors of adult substance abuse (Tennant et al., 1975). Therefore, the Results section indicates when the findings differ according to whether this study is included or omitted. The remaining six studies yielded a weighted mean of d = −.09. The 19 stud- ies that overlapped with Gershoff’s meta-analysis yielded a weighted mean effect size of d = −.23 in Gershoff, compared to d = −.10 according to our re- vised effect sizes. Both means indicated slightly detri- mental associations, but less so than in Gershoff’s overall meta-analysis.
The differences in effect sizes between the two meta-analyses are due to a variety of factors, two of which stand out. First, only 26% (5 of 19) of the Gershoff (2002) studies in this meta-analysis mea- sured overly severe physical punishment, compared to 65% (34 of 52) of studies with aggression- composite outcomes in her meta-analysis (Baumrind et al., 2002). Studies of overly severe physical punish- ment were less likely to investigate alternative dis- ciplinary tactics, probably because they emphasized the severity of physical punishment as an indica- tor of dysfunctional parent–child relations. In con- trast, studies of customary or conditional physical
34 Larzelere and Kuhn
punishment were more likely to investigate alterna- tive disciplinary tactics also, thus qualifying for this meta-analysis.
Second, only 3 of the 19 overlapping studies had similar effect sizes in both meta-analyses. Reasons for the discrepancies resulted from our decisions to include non-significant associations in calculating ef- fect size averages (three studies), to include all rele- vant associations (two studies), to use pre-post gain scores in the Roberts’s series of studies (three over- lapping studies), to use the best statistics available (longitudinal associations based on multiple sources of information specific to physical punishment rather than other associations in three studies), to use a different outcome variable (one study), and to use statistics that were similar for alternative tactics (one additional study). We could not replicate Gershoff’s estimates in three other studies in which her effect sizes indicated more beneficial outcomes of physical punishment than our estimates. Overall, the mean absolute difference in effect size estimates was .76 for the 19 overlapping studies (unweighted mean). In two extreme cases, our effect size was 2.17 more favorable than Gershoff’s d for Day and Roberts (1983), whereas our effect size was 4.42 less favorable than Gershoff’s d for Larzelere et al. (1996). Table I indicates how the effect size estimates in the current meta-analysis differ from Gershoff’s estimates.
Effect Size Estimates for Three Sets of Studies
The second part of this appendix specifies how effect sizes were estimated for three sets of stud- ies. These were selected because (1) they included the greatest discrepancies between the effect sizes estimates in this meta-analysis compared to Ger- shoff’s (2002) meta-analysis, (2) the estimating meth- ods could not be easily summarized in one phrase in Table I, (3) and the studies are particularly relevant for conditional spanking.
In the series of studies by Roberts and his col- leagues, our effect sizes for each tactic used to en- force time-out were based on two outcomes: compli- ance to parental clean-up commands (also used by Gershoff) and compliance with time-out in the clinic session. Improvement in compliance to parental commands was estimated with a pre-to-post treat- ment effect size. The child-determined release con- dition (Bean & Roberts, 1981; Roberts & Powers, 1990) was treated as the control condition. There- fore, its mean effect size was subtracted from the
effect size for each other tactic, to yield that tactic’s additional gain in compliance beyond that produced by a child-determined release from time-out. In three of the four studies, tactics could also be compared on post-treatment measures of the child’s compliance with time-out. The final effect size for each tactic was the difference between it and the mean effect size of the child-determined release condition. The pur- pose of these estimation methods were (1) to esti- mate the effect sizes using similar methods for each tactic, (2) to have the same standards of comparison for each study by Roberts and his colleagues, and (3) to expand the outcome variables to include compli- ance with time out as well as compliance to parental commands.
Ritchie’s (1999) study was particularly relevant for comparing the ability of eight disciplinary tactics to immediately stop several distinct types of noncom- pliance. Of particular interest for this meta-analysis were comparisons between the conditional proba- bility that each type of noncompliance immediately preceded each disciplinary tactic and the conditional probability that it immediately followed that same disciplinary tactic. For example, we wanted to com- pare the probability that defiance occurred immedi- ately before a spanking with the probability that de- fiance would be occurring immediately after it was used. The conditional probabilities subsequent to each tactic were given by Ritchie (1999). However, the immediately preceding conditional probabilities had to be estimated from information in the article (Larzelere, 2002). These estimates are only approx- imate and could be off by as much as 28%, based on the fact that the preceding conditional probabil- ities summed to totals ranging from 83 to 128%. The preceding conditional probabilities were adjusted for the discrepancy of these totals from 100%, but this merely averaged the estimation errors across the six types of noncompliance. Nonetheless, we estimated effect sizes based on immediate changes in the con- ditional probabilities of two types of noncompliance. The first type of noncompliance was defiance. The ef- fect size for spanking and changes in defiance were categorized as conditional physical punishment. The second type of noncompliance was called milder non- compliance, based on the average effect size for pas- sive and physical noncompliance. The contrast in ef- fect sizes for different tactics varied quite a bit for defiance vs. milder noncompliance.
We based the effect sizes from Larzelere et al. (1996) from deviation delays, because they provided stronger causal evidence than simple delays until a
Physical Punishment vs. Alternative Tactics 35
misbehavior recurrence (see rationale in Larzelere, 1996). Further, each effect size was calculated rel- ative to the effect size for “other” disciplinary re- sponses, which did not include reasoning or any type of punishment, and was therefore treated as the control condition. When effect sizes were averaged across categories (e.g., for both fighting and noncom- pliance incidents), they were weighted by the number of discipline incidents in each category.
ACKNOWLEDGMENT
This research was supported in part by Grant 1R03HD044679 from the National Institute of Child Health and Human Development. A previous version of this article was presented at the biennial convention of the Society for Research in Child Development, Tampa, FL, April 2003.
REFERENCES
(References marked with an asterisk indicate studies included in the meta-analysis.)
∗Aronfreed, J. (1961). The nature, variety, and social patterning of moral responses to transgression. Journal of Abnormal and Social Psychology, 63, 223–241.
Aronfreed, J. (1968). Aversive control of socialization. In W. J. Arnold (Ed.), Nebraska Symposium on Motivation (Vol. 16, pp. 271–320). Lincoln: University of Nebraska Press.
Axelrod, S., & Apsche, J. (Eds.). (1983). The effects of punishment on human behavior. New York: Academic Press.
Bauman, L. J., & Friedman, S. B. (1998). Corporal punishment. Pediatric Clinics of North America, 45, 403–414.
Baumrind, D. (1973). The development of instrumental compe- tence through socialization. In A. D. Pick (Ed.), Minnesota Symposia on Child Psychology (Vol. 7, pp. 3–46). Minneapo- lis: University of Minnesota Press.
Baumrind, D. (1983). Specious causal attributions in the social sci- ences: The reformulated stepping-stone theory of heroin use as exemplar. Journal of Personality and Social Psychology, 45, 1289–1298.
Baumrind, D. (1991). The influence of parenting style on adoles- cent competence and substance use. Journal of Early Adoles- cence, 11, 56–95.
Baumrind, D., Larzelere, R. E., & Cowan, P. A. (2002). Ordinary physical punishment: Is it harmful? Comment on Gershoff (2002). Psychological Bulletin, 128, 580–589.
∗Bean, A. W., & Roberts, M. W. (1981). The effect of time-out re- lease contingencies on changes in child noncompliance. Jour- nal of Abnormal Child Psychology, 9, 95–105.
Bee, H. (1998). Lifespan development (2nd ed.). Reading, MA: Addison-Wesley.
Bell, R. Q., & Harper, L. V. (1977). Child effects on adults. Hills- dale, NJ: Erlbaum.
Benjet, C., & Kazdin, A. E. (2003). Spanking children: The con- troversies, findings, and new directions. Clinical Psychology Review, 23, 197–224.
Berger, K. S., & Thompson, R. A. (1995). The developing per- son through childhood and adolescence (4th ed.). New York: Worth.
Bornstein, M. H., & Lamb, M. E. (Eds.). (1988). Developmental psychology: An advanced textbook (2nd ed.). Hillsdale, NJ: Erlbaum.
Brezina, T. (1999). Teenage violence toward parents as an adapta- tion to family strain: Evidence from a national survey of male adolescents. Youth & Society, 30, 416–444.
∗Burton, R. V., Maccoby, E. E., & Allinsmith, W. (1961). An- tecedents of resistance to temptation in four-year-old chil- dren. Child Development, 32, 689–710.
Campbell, D. T., & Boruch, R. F. (1975). Making the case for ran- domized assignment to treatments by considering the alter- natives: Six ways in which quasi-experimental evaluations in compensatory education tend to underestimate effects. In C. A. Bennett & A. A. Lumsdaine (Eds.), Evaluation and ex- periment: Some critical issues in assessing social programs (pp. 195–296). New York: Academic Press.
Campbell, D. T., & Erlebacher, A. E. (1970). How regression arti- facts in quasi-experimental evaluations can mistakenly make compensatory education look harmful. In J. Hellmuth (Ed.), Disadvantaged child: Vol. 3. Compensatory education: A na- tional debate (pp. 185–210). New York: Brunner/Mazel.
∗Chapman, M., & Zahn-Waxler, C. (1982). Young children’s com- pliance and noncompliance to parental discipline in a natural setting. International Journal of Behavioral Development, 5, 81–94.
Cohen, J. (1988). Statistical power analysis for the behavioral sci- ences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
∗Coopersmith, S. (1967). The antecedents of self-esteem. San Fran- cisco, CA: Freeman.
∗Crowne, D. P., Conn, L. K., Marlowe, D., & Edwards, C. N. (1969). Some developmental antecedents of level of aspira- tion. Journal of Personality, 37, 73–92.
Damer, T. E. (1980). Attacking faulty reasoning. Belmont, CA: Wadsworth.
∗Day, D. E., & Roberts, M. W. (1983). An analysis of the physical punishment component of a parent training program. Journal of Abnormal Child Psychology, 11, 141–152.
EPOCH-Worldwide. (2004). Legal reform: Corporal punishment of children in the family. Retrieved on October 27, 2004 from http://www.stophitting.com/laws/legalReform.php.
Etaugh, C., & Rathus, S. A. (1995). The world of children. Or- lando, FL: Harcourt Brace.
Eyberg, S. (1993). The spank back-up in time-out with preschool children. Unpublished manuscript, University of Florida, Gainesville.
Eysenck, H. (1993). Letter to the editor: Hitting the right cause. The Psychologist, 6, 392.
Friedman, S. B., & Schonberg, S. K. (1996a). Consensus state- ments. Pediatrics, 98, 852–853.
Friedman, S. B., & Schonberg, S. K. (1996b). [Personal statement]. Pediatrics, 98, 857–858.
Gershoff, E. T. (2002). Corporal punishment by parents and as- sociated child behaviors and experiences: A meta-analytic and theoretical review. Psychological Bulletin, 128, 539– 579.
Glass, G. V., McGaw, B., & Smith, M. L. (1981). Meta-analysis in social research. Beverly Hills, CA: Sage.
Graziano, A. M., Hamblen, L., & Plante, W. A. (1996). Subabusive violence in child rearing in middle-class American families. Pediatrics, 98, 845–848.
∗Grinder, R. E. (1962). Parental childrearing practices, conscience, and resistance to temptation of sixth grade children. Child De- velopment, 33, 803–820.
Grogan-Kaylor, A. (2004). The effect of corporal punishment on antisocial behavior in children. Social Work Research, 28, 153–162.
36 Larzelere and Kuhn
Grolnick, W. S., Deci, E. L., & Ryan, R. M. (1997). Internaliza- tion within the family: The self-determination theory perspec- tive. In J. E. Grusec & L. Kuczynski (Eds.), Parenting and children’s internalization of values (pp. 135–161). New York: Wiley.
Grusec, J. (1997). A history of research on parenting strategies and children’s internalization of values. In J. E. Grusec & L. Kuczynski (Eds.), Parenting and children’s internalization of values (pp. 3–22). New York: Wiley.
Grusec, J. E., & Goodnow, J. J. (1994). Impact of parental dis- cipline methods on the child’s internalization of values: A reconceptualization of current points of view. Developmental Psychology, 30, 4–19.
Gunnoe, M. L., & Mariner, C. L. (1997). Toward a developmental- contextual model of the effects of parental spanking on children’s aggression. Archives of Pediatrics and Adolescent Medicine, 151, 768–775.
∗Hall, E. C. (1994). A correlational analysis of parental conflict resolution practices and 4- and 5-year-old children’s interper- sonal problem-solving skills and verbal abilities in a preschool setting (Doctoral dissertation, University of San Francisco, 1994). Dissertation Abstracts International, 55(12A), 3785.
Harris, J. R. (1998). The nurture assumption. New York: Free Press.
Hedges, L. V. (1994). Fixed effects models. In H. Cooper & L. V. Hedges (Eds.), The handbook of research synthesis (pp. 285– 300). New York: Russell Sage Foundation.
Hoffman, M. L. (1970). Conscience, personality, and socialization techniques. Human Development, 13, 90–126.
Hoffman, M. L. (1977). Moral internalization: Current theory and research. In L. Berkowitz (Ed.), Advances in experimental so- cial psychology (Vol. 10, pp. 85–133). New York: Academic Press.
Holden, G. W. (1997). Parents and the dynamics of child rearing. Boulder, CO: Westview Press.
Ispa, J. M., & Halgunseth, L. C. (2004). Talking about corpo- ral punishment: Nine low-income African American mothers’ perspectives. Early Childhood Research Quarterly, 19, 463– 484.
Johnson, B. T. (1989). DSTAT: Software for the meta-analytic re- view of research literatures. Hillsdale, NJ: Erlbaum.
Kadushin, A., & Martin, J. A. (1981). Child abuse: An interactional event. New York: Columbia University Press.
Kochanska, G., Padavich, D. L., & Koenig, A. L. (1996). Chil- dren’s narratives about hypothetical moral dilemmas and ob- jective measures of their conscience: Mutual relations and so- cialization antecedents. Child Development, 67, 1420–1436.
Larzelere, R. E. (2000). Child outcomes of nonabusive and cus- tomary physical punishment by parents: An updated litera- ture review. Clinical Child and Family Psychology Review, 3, 199–221.
Larzelere, R. E. (2001). Combining love and limits in authoritative parenting. In J. C. Westman (Ed.), Parenthood in America (pp. 81–89). Madison: University of Wisconsin Press.
Larzelere, R. E. (2002). The effectiveness of alternative disciplinary tactics in reducing various types of noncompliance within ex- tended discipline episodes. Unpublished manuscript, Univer- sity of Nebraska Medical Center, Omaha.
Larzelere, R. E. (2004). Weighted analyses of the Canadian preva- lence of physical punishment by children’s age from Cycle 1 of the National Longitudinal Survey of Children and Youth. Unpublished raw data.
Larzelere, R. E., Baumrind, D., & Polite, K. (1998). Two emerging perspectives of parental spanking from two 1996 conferences. Archives of Pediatrics and Adolescent Medicine, 152, 303–305.
Larzelere, R. E., & Johnson, B. (1999). Evaluation of the effects of Sweden’s spanking ban on physical child abuse rates: A lit- erature review. Psychological Reports, 85, 381–392.
∗Larzelere, R. E., Klein, M., Schumm, W. R., & Alibrando, S. A., Jr. (1989). Relations of spanking and other parenting char- acteristics to self-esteem and perceived fairness of parental discipline. Psychological Reports, 64, 1140–1142.
Larzelere, R. E., Kuhn, B. R., & Johnson, B. (2004). The interven- tion selection bias: An underrecognized confound in interven- tion research. Psychological Bulletin, 130, 289–303.
∗Larzelere, R. E., Sather, P. R., Schneider, W. N., Larson, D. B., & Pike, P. L. (1998). Punishment enhances reasoning’s effec- tiveness as a disciplinary response to toddlers. Journal of Mar- riage and the Family, 60, 388–403.
∗Larzelere, R. E., Schneider, W. N., Larson, D. B., & Pike, P. L. (1996). The effects of discipline responses in delaying toddler misbehavior recurrences. Child & Family Behavior Therapy, 18(3), 35–57.
∗Larzelere, R. E., & Smith, G. L. (2000, August). Controlled longi- tudinal effects of five disciplinary tactics on antisocial behavior. Paper presented at the meeting of the American Psychologi- cal Association, Washington, DC.
Lipsey, M. W., & Wilson, D. B. (1993). The efficacy of psycho- logical, educational, and behavioral treatment: Confirmation from meta-analysis. American Psychologist, 48, 1181–1209.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage.
∗Lytton, H. (1977). Correlates of compliance and the rudiments of conscience in two-year-old boys. Canadian Journal of Be- havioural Science, 9, 242–257.
Matson, J. L., & Taras, M. (1989). A 20 year review of punish- ment and alternative methods to treat problem behaviors of developmentally delayed persons. Research in Developmental Disabilities, 10, 85–104.
∗McClelland, D. C., & Pilon, D. A. (1983). Sources of adult mo- tives in patterns of parent behavior in early childhood. Jour- nal of Personality and Social Psychology, 44, 564–574.
McNeil, C. B., Clemens-Mowrer, L., Gurwitch, R. H., & Funder- burk, B. W. (1994). Assessment of a new procedure to pre- vent timeout escape in preschoolers. Child & Family Behavior Therapy, 16(3), 27–35.
∗Minton, C., Kagan, J., & Levine, J. A. (1971). Maternal control and obedience in the two-year-old. Child Development, 42, 1873–1894.
National Institutes of Health. (1991). Treatment of destructive be- haviors in persons with developmental disabilities (NIH Publi- cation No. 91-2410). Bethesda, MD: Author.
Newsom, C., Favell, J. E., & Rincover, A. (1983). Side effects of punishment. In S. Axelrod & J. Apsche (Eds.), The effects of punishment on human behavior (pp. 285–316). New York: Academic Press.
Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia Press.
∗Ritchie, K. L. (1999). Maternal behaviors and cognitions during discipline episodes: A comparison of power bouts and single acts of noncompliance. Developmental Psychology, 35, 580– 589.
Roberts, M. W. (1984). An attempt to reduce timeout resistance in young children. Behavior Therapy, 15, 210–216.
∗Roberts, M. W. (1988). Enforcing chair timeouts with room time- outs. Behavior Modification, 12, 353–370.
∗Roberts, M. W., & Powers, S. W. (1990). Adjusting chair timeout enforcement procedures for oppositional children. Behavior Therapy, 21, 257–271.
Rothman, K. J., & Greenland, S. (1998). Modern epidemiology (2nd ed.). Philadelphia: Lippincott-Raven.
∗Sears, R. R. (1961). Relation of early socialization experiences to aggression in middle childhood. Journal of Abnormal and Social Psychology, 63, 466–492.
∗Sears, R. R., Maccoby, E. E., & Levin, H. (1957). Patterns of child-rearing. New York: Harper & Row.
Physical Punishment vs. Alternative Tactics 37
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experi- mental and quasi-experimental designs for generalized causal inference. Boston: Houghton Mifflin.
Shadish, W. R., & Haddock, C. K. (1994). Combining estimates of effect size. In H. Cooper & L. V. Hedges (Eds.), Handbook of research synthesis (pp. 261–281). New York: Russell Sage Foundation.
Simons, R. L., Lin, K.-H., & Gordon, L. C. (1998). Socialization in the family of origin and male dating violence: A prospec- tive study. Journal of Marriage and the Family, 60, 467– 478.
Smith, D. (2002). Journal article reignites debate over corporal punishment. Monitor on Psychology, 33(8), 14.
Statistics Sweden. (1996). Spanking and other forms of phys- ical punishment (Demography, the Family, and Children 1996:1.2). Stockholm: Author.
Stattin, H., Janson, H., Klackenberg-Larsson, I., & Magnusson, D. (1995). Corporal punishment in everyday life: An intergener- ational perspective. In J. McCord (Ed.), Coercion and pun- ishment in long-term perspectives (pp. 315–347). Cambridge, England: Cambridge University Press.
Straus, M. A. (2001). Beating the devil out of them: Corporal pun- ishment in American families and its effects on children (2nd ed.). New Brunswick, NJ: Transaction.
∗Straus, M. A., & Mouradian, V. E. (1998). Impulsive corporal punishment by mothers and antisocial behavior and impul- siveness of children. Behavioral Sciences and the Law, 16, 353–374.
Straus, M. A., & Paschall, M. J. (1998, August 1). Corporal pun- ishment by mothers and child’s cognitive development: A lon- gitudinal study. Paper presented at the 14th World Congress of Sociology, Montreal, Canada.
Straus, M. A., & Stewart, J. H. (1999). Corporal punishment by American parents: National data on prevalence, chronicity, severity, and duration, in relation to child and family char- acteristics. Clinical Child and Family Psychology Review, 2, 55–70.
Straus, M. A., Sugarman, D. B., & Giles-Sims, J. (1997). Spank- ing by parents and subsequent antisocial behavior of children. Archives of Pediatrics and Adolescent Medicine, 151, 761–767.
∗Tennant, F. S., Jr., Detels, R., & Clark, V. (1975). Some childhood antecedents of drug and alcohol abuse. American Journal of Epidemiology, 102, 377–385.
Turner, H. A., & Muller, P. A. (2004). Long-term effects of child corporal punishment on depressive symptoms in young adults: Potential moderators and mediators. Journal of Fam- ily Issues, 25, 761–782.
Walters, G. C., & Grusec, J. E. (1977). Punishment. San Francisco: Freeman.
∗Watson, D. G. (1989). Parenting styles and child behavior: A study of retrospective reports from parents of 2500 high school students (Doctoral dissertation, State University of New York at Buffalo, 1989). Dissertation Abstracts Interna- tional, 50(7B), 3181.
Webster-Stratton, C. (1990). Enhancing the effectiveness of self- administered videotape parent training for families with conduct-problem children. Journal of Abnormal Child Psy- chology, 18, 479–492.
∗Yarrow, M. R., Campbell, J. D., & Burton, R. V. (1968). Child rearing: An inquiry into research and methods. San Francisco: Jossey-Bass.
∗Zahn-Waxler, C., Radke-Yarrow, M., & King, R. (1979). Child rearing and children’s prosocial initiations toward victims of distress. Child Development, 50, 319–330.
