SPSS Building a Basic Data Set

vwccspt12
TypesofVariablesandLevelsofMeasurement1.pdf

W

A L

D E

N U

N I

V E

R S

I T

Y

P

U B

H 6

0 3

2 /

8 0

3 2

S

P S

S R

E V

E A

L E

D

P

U B

H 6

0 3

3 /

8 0

3 3

I

N T

E R

P R

E T

A T

I O

N &

A P

P O

F D

A T

A

T yp

es of V ariables &

L

evels of M easu

rem en

t

T yp

es of V ariables

y A

variable is an y factor, trait, con

d ition

, or d isease

th at exists in

d ifferin

g am ou

n ts or typ

es. y

E xp

erim en

tal an d

observation al stu

d ies con

sist of several typ

es of variables. y

A ll stu

d ies in

volve at least on e in

d ep

en d

en t variable

(IV ) an

d on

e d ep

en d

en t variable (D

V ).

{

G et acqu

ain ted

w ith

id en

tifyin g th

e IV an

d D

V . It’s th

e first step

in d

evelop in

g a research qu

estion .

D ep

en d

en t V

ariable (D V

)

y D

isease or h ealth

ou tcom

e of in terest

y P

resu m

ed effect of a stu

d y

y Syn

on ym

s: O u

tcom e or C

riterion variable

y E

xam p

les: {

C an

cer {

D iabetes

{ H

IV statu

s {

C H

D m

ortality {

D en

gu e in

fection

In d

ep en

d en

t V ariable (IV

)

y P

red ictor or exp

osu re of in

terest y

P resu

m ed

cau se of a stu

d y

y Syn

on ym

s: E xp

erim en

tal, P red

ictor, or M an

ip u

lated

variable y

E xam

p les:

{ A

ge {

R ace

{ In

com e

{ B

M I

{ D

ru g U

se {

C on

d om

U se

C on

fou n

d in

g V ariable

y A

variable th at obscu

res th e effects betw

een th

e IV

an d

D V

y

A variable th

at th e research

er failed to con

trol or elim

in ate

y Syn

on ym

s: C ovariate, T

h ird

variable, M ed

iatin g

variable y

E xam

p les:

{ C

igarette sm okin

g con fou

n d

s th e effect betw

een d

rin kin

g coffee (IV

) an d

h eart d

isease (D V

) {

D iet, age, an

d gen

d er con

fou n

d th

e effect betw een

lack of exercise (IV

) an d

w eigh

t gain (D

V )

L evels of M

easu rem

en t

y E

ach variable can

be classified by its level of

m easu

rem en

t. y

W e w

ill classify accord in

g to term s u

sed in

th e

G erstm

an (20

16 ) textbook. H

ere th ey are listed

in

ord er of in

creasin g p

recision :

{ C

ategorical {

O rd

in al

{ Q

u an

titative

{ It is im

p ortan

t to u n

d erstan

d th

e d ifferen

t levels of m

easu rem

en t as th

ey h elp

d eterm

in e th

e statistical test th at

sh ou

ld be carried

ou t.

C ategorical

y C

lassify observation s in

to n am

ed categories, w

ith n

o in

trin sic ord

erin g of th

e categories y

C ategories are often

assign ed

n u

m erical valu

es as labels (eg., 0

= N

o, 1= Y

es) y

Syn on

ym s: N

om in

al y

C ategorical variables w

ith 2 levels are also called

B

in ary or D

ich otom

ou s

y E

xam p

les: {

G en

d er: M

ale/F em

ale {

R ace: B

lack/W h

ite/O th

er {

M ortality statu

s: D ead

/A live

O rd

in al

y C

ategories th at can

be p u

t in ran

k ord er

y E

xam p

les: {

Stage of can cer (I, II, III, IV

) {

Sm okin

g frequ en

cy (n on

-sm oker, ligh

t sm oker, m

od erate

sm oker, h

eavy sm oker)

{ Y

ear in college (fresh

m an

, sop h

om ore, ju

n ior, sen

ior)

Q u

an titative

y T

ru e n

u m

erical valu es th

at can be p

u t on

a n u

m ber

lin e

y Syn

on ym

s: C on

tin u

ou s, Scale, an

d In

terval y

E xam

p les:

{ A

ge (years) {

B M

I (kg/m 3)

{ Seru

m ch

olesterol (m g/d

L )