Article Component Papers (LEADERSHIP TRAITS AND SKILLS)

Journal of Applied Psychology 1962, Vol. 46, No. 4, 291-295

SELF-CONFIDENCE AND LEADERSHIP

DAVID KIPNIS AND WILLIAM P . LANE i

United States Naval Personnel Research Activity, Washington, D. C.

The object of this study was to examine the relationship between lack of confidence in one's leadership ability and reliance upon passive leadership techniques to cope with supervisory problems. 77 Navy petty officers were given a questionnaire containing 20 supervisory problems and were asked to evaluate the desirability of each of 5 ways of solving each problem. Ss also evaluated how satisfied they were with their leadership abilities. Principal findings were: (a) Ss were highly consistent in the extent to which they endorsed each of 5 approaches to correcting performance, (b) there was a correlation of .52 between endorsing the use of administrative procedures to solve the problem and informally asking a superior to solve the problem, and (c) Ss who lacked confidence in their leadership abilities were significantly less willing to hold face-to-face discussions with subordinates and significantly more often endorsed both referring the subordinate to a superior and relying upon the use of administrative rules to solve the supervisory problems.

As Gibb (1961) has pointed out, a rather important component of effective leadership is how well the leader uses the powers in- herent in his role for directing followers. This observation would seem to be specially true in business and military organizations where supervisors are frequently chosen for their technical rather than leadership skills. Be- cause of this selection on other than leader- ship skills, many supervisors lack the charis- matic qualities needed to develop personal loyalties, and of necessity, must rely upon the application of the powers provided them by the organization to successfully direct subordinates' performance. Thus it is of in- terest in hierarchical structures to investigate the repertoire of actions available to leaders and how this repertoire is used.

Petty officers (POs) are the Navy's first line supervisors. In a previous study (Kipnis, Lane, & Frankfurt, 1961) the repertoire of actions available to these POs for correcting subordinates' behavior was investigated by analyzing POs descriptions of how they cor- rected subordinates' performance. Among the results was the finding that chief petty of- ficers (CPOs), who are the most senior en- listed supervisors, used direct face-to-face actions to correct subordinates more fre-

1 The views expressed here do not necessarily represent those of the United States Navy.

The authors wish to thank Albert S. Glickman for his helpful suggestions and criticisms.

quently than did second class petty officers (P02s). The P02s relied on the more pas- sive actions of either asking the help of a superior in dealing with the subordinate or placed the subordinate on report. Placing a subordinate on report is a formal disciplinary procedure that may lead to judicial action against the subordinate. At a minimum, it requires the commanding officer to decide how to deal with the subordinate.

There was, of course, considerable overlap in the corrective actions taken by CPOs and P02s. Not all P02s placed a man on report or referred him unofficially to a superior, and not all CPOs took direct face-to-face correc- tive actions. This finding suggested that it might not be simply variations in authority and power which accounted for the differences between CPOs and P02s, although clearly the factors involved were associated with these differences.

The present study examined whether a PO's confidence in his leadership ability was related to his reliance upon passive or non- passive leadership techniques to cope with supervisory problems. Since self-confidence tends to be correlated with amount of experi- ence and authority status, one would expect junior POs to be less confident than CPOs in their leadership capabilities. And since there is some experimental evidence (Hochbaum, 1954) that low degrees of self-confidence lead to more reliance upon the opinions of others,

291

292 DAVID K I P N I S AND WILLIAM P. LANE

it seems reasonable that junior POs' inclina- tions to rely upon other persons, officially or unofficially, for help in solving their super- visory problems may be attributed to this variable.

PROCEDURE

Correction of Performance Questionnaire. The Cor- rection of Performance Questionnaire (COP) was based upon incidents gathered in earlier research (Kipnis et al., 1961) in which POs described how they corrected subordinates' performance. Twenty incidents which described substandard behaviors of enlisted men were used. Following each incident, five alternatives were offered, each describing a different way of handling the problem.

Alternative "A" advocated talking to the sub- ordinate to find out why he acted as he did; Alternative " B " advocated reprimanding the m a n ; " C " advocated the use of sanctions such as extra instruction, denial of privileges, and inspection; " D " advocated informal referral of the problem to a superior for action; " E " advocated officially placing the subordinate on report.

The POs frequently use more than one corrective action in dealing with a subordinate's performance. Hence, it would have been unrealistic to restrict the number of alternatives POs could choose to handle the problem. Accordingly, the instructions required respondents to rate each alternative on a five-point scale, ranging from "A Highly Desirable Action" (HD) to a "Very Undesirable Action" (VU). A sample item follows:

1. A seaman refused to take suggestions on how work should be performed. When given a direct order, the man became angry and answered back, although he did not refuse to do the work.

a. Talk with the man to find out the cause of his actions.

H D D ? U VU

b. Warn the man that he H D D ? U VU will be reported for disre- spect.

c. Deprive the man of H D D ? U VU special privileges for a week.

d. Ask the Division Officer H D D ? U VU to handle the man.

e. Place the man on report H D D ? U VU for disrespect and failure to follow instructions.

Weights of 5, 4, 3, 2, 1 were assigned to the scale (with H D = 5 ) . Five scores were derived from the questionnaire, each with a possible range of 20 to 100: Diagnostic Talk, Reprimand, Sanctions, Re-

ferral, and Report.

TABLE 1

S P L I T - H A L F RELIABILITIES OF THE C O P

(N = 77)

COP scores Reliability

Diagnostic Talk .91 Reprimand .82 Sanctions .84 Referral .89 Report .90

Self-Confidence Questionnaire. Respondents were also asked to indicate how satisfied they were with their own performance in each of seven leadership skills:

1. Supervise your work crew's technical work. 2. Help your men prepare for advancement. 3. Correct lax or unmilitary behavior in your men. 4. Get your men to work productively on hard

duty. 5. Criticize constructively poor performance of

your men. 6. Develop favorable attitudes toward the Navy

among your men. 7. Get your decisions backed up by your superior.

The five-point scale ranged from "Very Satisfied," weighted 5, to "Not Satisfied at All," with a weight of 1. A Self-Confidence score for each man was obtained by summing weights over all seven items for a possible range of 7 to 35.

Sample. The questionnaires were administered prior to training at three Navy Petty Officer Leader- ship Schools. Seventy-seven usable questionnaires were obtained from 27 CPOs, 44 POls, and 6 P 0 2 s . The POls are intermediate in supervisory rank be- tween CPOs and P 0 2 s .

The sample was highly selected, since the policy of the Navy is to assign only POs who show above average leadership abilities to these schools. I t is expected that the POs, upon completion of training, will help set up leadership training courses at their next duty stations.

Questionnaire Reliability. Table 1 gives the cor- rected split-half reliabilities of the five scores of the COP. These ranged from .82 to .91. The previous research (Kipnis et al., 1961) found POs were con- sistent over several leadership problems in whether or not they would talk to a subordinate about his poor performance, would refer him to a superior, or would place him on report. The reliabilities reported in Table 1 tend to confirm those findings of in- dividual consistency in choice of leadership actions.

The corrected split-half reliability of the Self- Confidence scale was .81.

RESULTS

The intercorrelations among the five scores of the COP are shown in Table 2. There was

SELF-CONFIDENCE AND LEADERSHIP 293

TABLE 2

INTERCORRELATIONS BETWEEN THE

(N = 77)

COP SCORES DISTRIBUTION

TABLE 3

OF SELF-CONFIDENCE SCORES

(By rank)

COP scores 1 2 3 4 5

Self-Confidence scoi

Rank

1. Diagnostic Talk — .07 - . 0 6 2. Reprimand — .44** 3. Sanctions — 4. Referral

- . 0 1 .28* .28*

- . 2 6 * .16 .19 .52**

Self-Confidence scoi

Chiefs (N = 27)

res %

P O l s + P 0 2 s (N = 50)

%

5. Report

- . 0 1 .28* .28*

- . 2 6 * .16 .19 .52**

High third Middle third Low third

41 33 26

26

* £ < . 0 5 . ** p <.01.

High third Middle third Low third

41 33 26

28 46

a fair degree of independence among the five scores, given that they were derived from the same instrument. The highest correlation was between the Referral and Report scores (r = .52, p < .01), suggesting that informally asking for help from a supervisor as well as formally

invoking administrative rules to solve one's problems may share a common psychological basis. It is also interesting to note that there was a negative correlation between Diagnostic Talk and Report (r = - .26, p < .05), sug-

TABLE 4

M E A N S AND SUMMARY OF ANALYSIS OF VARIANCE OF THE F I V E COP SCORES

Mean COP scores Rank

Analysis of variance

Self-Confidence Mean COP scores

Rank Source df MS F

Diagnostic Talk

Average-high Low

Chief 89.65 81.00

P O l + P 0 2 86.52 82.56

Reprimand

Rank Confidence C X R Error

1 1 1

73

.61 39.74 5.51 7.47

.08 5.51* .74

Average-high Low

Chief 73.30 76.43

P O l + P 0 2 76.11 75.30

Sanctions

Rank Confidence C X R Error

1 1 1

73

1.35 2.20 2.39 5.24

.26

.42

.46

Average-high Low

Chief 69.60 71.43

P O l + P 0 2 69.00 71.74

Referral

Rank Confidence C X R Error

1 1 1

73

.03 5.36

.06 5.30

.01 1.01 .01

Average-high Low

Chief 47.55 54.00

P O l + P 0 2 47.51 52.65

Report

Rank Confidence C X R Error

1 1 1

73

1.42 29.09

.28 5.58

.25 5.21*

.05

Average-high Low

Chief 42.25 53.71

P O l + P 0 2 46.22 50.39

Rank Confidence C X R Error

1 1 1

73

.15 61.12 13.25 9.57

.02 6.38* 1.38

Note.—Average-high Self-Confidence: Chiefs = * p <.05.

20, POl + P 0 2 =27. Low Self-Confidence: Chiefs = 7, POl + P 0 2 = 2 3 .

294 DAVID K I P N I S AND WILLIAM P. LANE

gesting that the greater the willingness of the PO to actively discuss the problem with his subordinate, the less likely he was to rely on administrative rules to solve the problem.

Self-Confidence scores were divided into approximately equal thirds and the percent- ages of CPOs and POls plus P02s falling into each third are shown in Table 3.

As was expected, more CPOs expressed con- fidence in their leadership skills than did POls + P02s. Thus, 26% of the CPOs and 46% of the POls + P02s were classified in the low third of Self-Confidence scores.2 While the direction of the differences was consistent with our original hypothesis, a chi square test of these differences was not significant (x2 = 3.20, df = 2,p< .30). Since the POs in our sample were originally highly selected in terms of their leadership skills, it seems likely that in a less selected sample, the trends ap- parent here would reach customary levels of significance.

The relationship of Self-Confidence with the five COP scores was determined through analysis of variance. Since the earlier work (Kipnis et al., 1961) had found a relation- ship between supervisory status and leader- ship practices, a factorial design was used including both Self-Confidence scores (divided into upper two-thirds and lower one-third) and Rank (CPOs versus POls + P02s) as inde- pendent variables. To take into account the unequal numbers in the various cells, a cor- rection proposed by Walker and Lev (1953, pp. 381-382) was used.

Table 4 presents the results of the analysis. It can be seen that those POs classified as low in Self-Confidence were significantly less willing to hold diagnostic talks with the sub- ordinate (/>< .05), and significantly more often endorsed both referring the subordinate to a superior (p < .05) and placing the sub- ordinate on report (p < .05).

On the other hand it is apparent that, within the limited range of CPOs, POls, and P02s studied here, PO status alone had no specific effect on COP scores. The few CPOs who were categorized as low in Self-Confi-

2 While not analyzed separately because of the small numbers, it is worth noting that none of the 6 P02s were classified in the high third of Self- Confidence scores.

dence had the same pattern of COP scores as the more junior POs who were so classified.

DISCUSSION

The results indicate that doubts about one's leadership abilities lead to what might loosely be called a "buckpassing" approach to solving leadership problems. These results tie in nicely with the previous findings that junior POs exercised less direct leadership and placed more reliance upon unofficial referrals to superiors or upon placing a subordinate on report. The suggested relationship seems to be: inexperience contributes to lack of confi- dence, which in turn leads to reliance on others for solving leadership problems.

It would seem that effective leadership problem solving in a hierarchical organiza- tion has at least two components. First, the supervisor must be able to intelligently select the appropriate corrective actions from the repertoire of behaviors available to him. In the previous research (Kipnis et al., 1961), it was found that the supervisor system- atically varied the kinds of corrective actions used as the problems presented by the sub- ordinate varied. Secondly, the supervisor must be willing to take action to solve the prob- lem and not get in the habit of placing the problem in someone else's lap. The importance of this motivational variable has been sug- gested in several reports. For example, Havron and McGrath (1961) reported that effective leaders were characterized by a "willingness to act" when a problem situation arose. Gibb (1961) has done work on the ability of the leader to use the repertoire of behaviors he has at his command, and Clark, Spector, and Glickman (1960) identified an effective Navy CPO as one who made full use of his super- visory powers.

It is evident that there are other factors besides self-confidence and experience that are related to a willingness to take direct leader- ship actions. For example, span of control of the supervisor may be important. We found that POs who supervised large groups of men relied more on referral and report and less on direct supervision than did POs super- vising smaller groups.

The correlations between the Referral and Report scores suggests that unofficially asking

SELF-CONFIDENCE AND LEADERSHIP 295

for help and advice from a superior and reliance upon administrative regulations share common psychological elements. Both of these actions seem to be part of the bureaucratic buckpassing syndrome; serving as a means for the individual leader to avoid taking personal action and making difficult decisions.

While psychologically these two actions may share common elements, administratively the consequences of reliance upon regulations are quite different from asking for advice in- formally. For instance, it raises the question as to whether subordinates may be subjected to disciplinary action because they are by unhappy chance subject to the supervision of an inadequate leader. Invoking administrative regulations as a substitute for personal action also burdens higher executive levels with decisions that should have been made at a much lower echelon. Obviously all levels of an organization suffer to the extent that super- visors too frequently rely on this mode of supervision.

R E F E R E N C E S

CLARK, R. A., SPECTOR, A. J., & GLICKMAN, A. S.

The relationship between supervisory character- istics and attitudes of subordinates. Personnel PsychoL, 1960, 13, 301-316.

GIBB, J. R. Defense level and influence potential in small groups. In L. Petrullo and B. M. Bass (Eds.), Leadership and interpersonal behavior. New York: Holt, Rinehart & Winston, 1961. Pp. 66-81.

HAVRON, M. D., & M C G R A T H , J . E. The contribu-

tion of the leader to the effectiveness of small military groups. In L. Petrullo and B. M . Bass (Eds.), Leadership and interpersonal behavior. New York: Holt, Rinehart & Winston, 1961. Pp. 167-178.

HOCHBAUM, G. The relation between group mem- bers' self-confidence and their reactions to group pressures to uniformity. Amer. sociol. Rev., 1954, 19, 678-687.

K I P N I S , D., LANE, W. P., & FRANKFURT, L. Leader-

ship problems and practices of petty officers. USN Bur. Naval Personnel tech. Bull., 1961, No. 61-13.

WALKER, HELEN, & LEV, J . Statistical inference. New York: Holt, 1953.

(Received January 29, 1962)