I 
 
 mm HHii 
 
LIBRARY 
 
 OF THF. 
 
 U N IVERSITY 
 
 or 1LLI NOIS 
 
 I£6t 
 no. 6-\2 
 
The person charging this material is re- 
 sponsible for its return on or before the 
 Latest Date stamped below. 
 
 Theft, mutilation and underlining of books 
 are reasons for disciplinary action and may 
 result in dismissal from the University. 
 
 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN 
 
 L161— O-1096 
 
*b 
 
37 O 
 
 GROUP EFFECTIVENESS RESEARCH LABORATORY 
 Department of Psychology 
 University of Illinois 
 Urbana, Illinois 
 
 THE RELATION OF 
 
 OPEN HEARTH FOREMEN'S INTERPERSONAL PERCEPTIONS 
 
 TO STEEL PRODUCTION 
 
 WALTER A. CLEVEN and FRED E. FIEDLER 
 TECHNICAL REPORT NO. 11 
 
 A NOTE ON PSYCHOLOGICAL ATTRIBUTES 
 
 RELATED TO THE SCORE ASSUMED SIMILARITY 
 
 BETWEEN OPPOSITES (ASo) 
 
 EILEEN F. GOLB and FRED E. FIEDLER 
 TECHNICAL REPORT NO. 12 
 
 Studies performed under Contract N6-ori-07135 
 with the Office of Naval Research 
 
 PROJECT ON 
 SOCIAL PERCEPTION AND GROUP EFFECTIVENESS 
 
 THE LIBRARY UF THE 
 MAR y 
 
 OCTOBER, 1955 
 
 N 
 
GROUF EFFECTIVENESS RESEARCH LABORATORY 
 DEFARTMENT OF PSYCHOLOGY 
 UNIVERSITY OF ILLINOIS 
 URBANA, ILLINOIS 
 
 THE RELATION OF OPEN HEARTH FOREMEN'S INTERPERSONAL 
 PERCEPTIONS TO STEEL PRODUCTION 
 
 Walter A. Cleven and Fred E. Fiedler 
 
 Technical Report No. 11 
 
 Study performed under Contract N6-ori-07135 
 with the Office of Naval Research 
 
 Project on 
 Social Perception and Group Effectiveness 
 
 October, 1955 
 
 J M LIBRARY OF THE 
 
 MAR 2 ? 
 
Digitized by the Internet Archive 
 
 in 2012 with funding from 
 
 University of Illinois Urbana-Champaign 
 
 
 http://www.archive.org/details/relationofopenhe11clev 
 
S7© 
 
 The Relation of Open-Hearth Foremen's Interpersonal 
 Perceptions to Steel Production 
 
 Walter A. Cleven and Fred E. Fiedler 
 University of Illinois 
 
 This investigation is one of a series of studies on the relationship of 
 interpersonal perception and group effectiveness (2, 3), It was designed 
 to test hypotheses which grew out of earlier studies on military aircraft 
 and tank crews (3), The research was conducted in open-hearth shops of 
 a large steel company in which the personnel, in contrast to military crews, 
 is highly stable over time, and where carefully maintained production re- 
 cords are available. 
 
 Interpersonal perception is measured here by means of the score, 
 Assumed .Similarity between Opposites, or ASo. This score reflects the 
 extent to which the subject (S) differentiates between the man with whom 
 he can work best, and the man with whom he can work least well. We 
 infer that ASo is related to the psychological distance which S perceives 
 between himself and his co-workers. Supervisors who perceive much 
 similarity between their best- and least-liked co-workers (high ASo) are, 
 by this interpretation of ASo, more accepting, approachable individuals, 
 
 We are indebted to Messrs. J. H. Vohr, P. E. Thomas, E. C. 
 Sorrells, H. W. Erler, and G. H. Warnock. of the Gary Steel Works; and 
 to Mr. Dan Farrell and Mr. E. W. Kempton, of the United States Steel 
 ^Corporation, for their cooperation and support of this study. We are also 
 ^indebted to Drs. L. J. Cronbach, Eleanor P. Godfrey, Ross Stagner, and 
 \C . F. Wrigley for their contributions to the design and administration of 
 $he study, and to Mrs. Betty F. Mannheim who assisted with the analysis 
 ^"of the data. 
 
 i 
 
while supervisors who perceived little similarity between these workers 
 (low ASo) are presumably more critical and analytic in their work re- 
 lationships. 
 
 Previous studies of informal teams (2) and of Air Force bomber crews 
 and army tank crews (3) led to two hypotheses concerning the relation- 
 ship between leader ASo and group effectiveness. 
 
 Hypothesis I_. More effective groups have supervisors who 
 tend to perceive larger differences in the personalities of 
 their most and their least preferred co-workers (low ASo), 
 than do supervisors of less effective groups. 
 Hypothesis II. More effective groups have supervisors who 
 maintain a moderate amount of psychological distance between 
 themselves and their co-workers. This hypothesis consists, 
 operationally, of two parts: 
 
 Hypothesis Ha. More effective groups have supervisors with 
 low ASo only when the supervisor is well-liked and when he 
 in turn likes certain key personnel in the group. (As can be 
 seen, Hypothesis Ha is a special case of Hypothesis I.) 
 Hypothesis lib. More effective groups have supervisors with 
 high ASo when the supervisor is well-liked but does not like 
 certain key subordinates in the group. (This hypothesis limits 
 the generality of Hypothesis I.) 
 The two parts of the second hypothesis assume an interaction between 
 supervisor ASo and supervisor preferences for important subordinates 
 in the group. By this hypothesis, effective groups have either low ASo 
 supervisors who sociometrically prefer their key subordinates, or high 
 ASo supervisors who do not sociometrically choose their key subordinates. 
 
 ^Hypotheses I and II are postulated as valid only for tasks which require 
 "direction-giving leadership behavior" (3, p. 234). In contrast, the study 
 of military aircraft groups suggested the hypothesis that more effective 
 groups on tasks which require receptive leadership behavior have super- 
 visors with high ASo, regardless of the supervisor's preferences for key 
 subordinates. We had hoped to shed further light on this hypothesis in 
 the present study using incidence of accidents to subordinates as a cri- 
 terion, but this criterion proved unreliable. 
 
Hypothesis II in part overlaps with Hypothesis I. Our studies of in- 
 formal groups, i.e., basketball and surveying teams, lead us to expect 
 a negative correlation between the ASo of the group's informal leader 
 and his group's effectiveness. Our studies of formal, military groups, 
 as indicated by Hypothesis II, lead us to anticipate a more complex leader- 
 follower interaction. Specifically, we found a negative correlation between 
 the leader's ASo score and the criterion of effectiveness for groups in 
 which the accepted leader chose his keyman, and a positive correlation 
 between the two variables for groups in which the accepted leader did not 
 choose his keyman. Both Hypotheses I and II were considered in this 
 study of steel shops because our knowledge and understanding of group 
 processes seemed insufficient to judge, a priori , which hypothesis would 
 best fit the case of these relatively long-lived industrial work groups. 
 
 Procedure 
 
 Samp le. Management personnel in four open-hearth shops of a large 
 steel company participated in the study. These four shops are engaged in 
 similar operations, although equipment varies somewhat from shop to shop. 
 
 Each shop is operated on a 24-hour, seven-day-week basis with each 
 shift or "turn" working eight hours. Every turn has a full complement of 
 first- and second-line supervisors and their crews. Since one turn is off 
 duty in any one 24-hour period, each shop requires four turns. A total of 
 16 turns thus constitutes our sample. 
 
 Four supervisors are in charge of each turn: one General Foreman, 
 one Stock Foreman, one Fit Foreman, and one Senior Melter. The General 
 Foreman, along with the Stock and Pit Foremen, directs the supporting 
 operations of raw material assembly and final steel pouring. The Senior 
 Melter is in charge of steel manufacture. Depending on the number and 
 size of furnaces in the shop, the Senior Melter supervises one or two 
 
 3 
 All four turns within a shop are under the direction of a single shop 
 
 superintendent and his assistant superintendent. These men were not 
 
 tested, but only completed the performance ratings which are discussed 
 
 below, 
 

 • 
 
Junior Melters and their crews. In three of the four shops (or 12 of the 
 16 turns), the Senior Melter has two Junior Melters reporting to him. 
 
 Test instruments . Two test instruments were administered to all 
 available foremen and melters. The first requested each S to predict the 
 responses of two men: (a) the man with whom he can work best, and (b) the 
 man with whom he can work least well. These ratees could be anyone with 
 whom S had ever worked; S was not asked to specify their names. The test 
 consisted of 40 statements such as: "I tend to join many organizations ," 
 "I am often bored with people," and "I am generally regarded as optimistic." 
 Each item was answered on a six-point scale ranging from "definitely true" 
 to "definitely untrue." The similarity of these two predictions, computed 
 by the statistic D (1, 6), yields the index Assumed Similarity between 
 Opposites (ASo). 
 
 The second test was a sociometric questionnaire. Each S was asked to 
 nominate in order of his preference three men for each supervisory posi- 
 tion other than his own, These nominations could be made irrespective of 
 the nominee's position or the shop in which he was employed at the time of 
 testing. Nominations were made for three criterion situations: (a) "three 
 men who you think would be best able to help you improve the turn's safety 
 record," (b) "three men with whom you would want to work on a special 
 job which needs to be done very quickly," and (c) "three men whom you like 
 best personally." 
 
 Criterion 
 
 The index of group effectiveness is based on the time elapsed from one 
 "tap" (pouring of molten metal from the furnace) to the next tap on a par- 
 ticular furnace. For economic reasons, company officials regard short 
 "tap-to-tap time" as the most important production gcal. The primary im- 
 portance of this production goal is recognized and accepted by the foremen 
 as well as their subordinates. 
 
 The average tap-to-tap time is about ten hours; two turns are, therefore, 
 involved in preparing each batch of steel, or "heat," for tapping. However, 
 the tap-to-tap time scores are uniformly assigned to the Junior Melter and 
 crew in charge of the furnace at the time the tap is made, regardless of 
 
the length of time the shift has actually worked on the heat. This seems 
 justifiable because the last hours of the heat are regarded as more criti- 
 cal in the manufacturing process than the first few hours. In addition, 
 randomization takes place because the turns of Junior Melters and their 
 crews do not systematically follow one another or use the same equipment. 
 
 Using an analysis of variance of ranked data, we found significant dif- 
 ferences between shops in tap-to-tap time. Since these differences can be 
 attributed to different furnace capacities, the tap-to-tap time data were 
 standardized within shops by means of T scores to permit intershop com- 
 parison. 
 
 Criterion reliability . The reliability of tap-to-tap scores was based on 
 an analysis of 25,000 heats based on the three to 16 months period preced- 
 ing testing. We excluded the summer months on recommendation of com- 
 pany officials because of extensive personnel shifts due to vacation schedules, 
 An even-month vs. odd-month split-half procedure was employed. The es- 
 timated reliability of tap-to-tap time over the 16 turns is .82-, In order to 
 minimize the effects of long-range changes, e.g„, in personnel or company 
 policy, the criterion scores used below arc based on only a part of these 
 data, namely the three- to ten-month period immediately preceding testing. 
 
 Results 
 
 Table 1 presents the correlations between the average turn tap-to-tap 
 time and the ASo (Assumed Similarity betv/een Cpposites) of the General, 
 Fit and Stock Foremen, and Senior Melters. As the table shows, the corre- 
 lation between average turn tap-to-tap time and ASo is significant in the 
 case of Senior Melters and Pit Foremen, The correlation falls short of 
 an acceptable significance level for Stock Foremen, and is negligible for 
 
 General Foremen. The average ASo of the foremen and Senior Melter on 
 
 4 
 each turn is also significantly related to average turn tap-to-tap time. 
 
 4 
 The ASo of Junior Melters was not significantly related to their individual 
 
 tap-to-tap scores. No a priori hypothesis was formulated regarding this 
 correlation, since the relationship of group effectiveness to ASo of group 
 members other than the formal or informal leader has not been systemati- 
 cally examined. 
 
TABLE 1 
 
 Correlations (Rho) between ASo of Various Supervisors and Average 
 
 Turn Tap-to-Tap Time 
 
 Supervisor ASo 
 
 N 
 
 Rho 
 
 General Foreman 
 Stock Foreman 
 Pit Foreman 
 Senior Melter 
 Supervisor average 
 
 15 
 15 
 14 
 15 
 16 
 
 -.13 
 -.42 
 -.72 
 -.54 
 -.71 
 
 < .01 
 
 < .05 
 <L .01 
 
 *N varies due to missing data. 
 
 Di scuss ion of main fi ndings . Hypothesis I states that more effective 
 groups have supervisors with low ASo. The overlapping hypothesis, 
 Hypothesis II, states that group efficiency is related to supervisor ASo, 
 but that the direction of relationship is a function of the supervisor's 
 preferences for his key subordinates. The significant correlations of 
 Senior Melter and Pit Foreman ASo, as well as average supervisor ASo, 
 with average turn tap-to-tap may be interpreted as supporting Hypothesis 
 I rather than Hypothes ; s II. It is unlikely that the distributions of points 
 comprising these significant correlations would break down (on the basis 
 of sociometric analysis) into two subdistributions with opposite slopes, 
 as the interaction hypothesis requires. We shall return to this point be- 
 low. 
 
 Of particular interest is the high correlation between mean turn tap 
 time and ASo of Pit Foremen, On the surface, the melter s appear to 
 determine tap time, since it is their decision as to when the heat is res.dy 
 to be tapped. The results suggest, however, that variance in tap-to-tap 
 time may also be a function of the Pit Foreman, although, conversely, 
 the ASo of these supervisors may be a function of the turn efficiency or 
 of some related variable. On the other hand, the low correlation in the 
 
case of the General Foremen may indicate that these men have the least 
 influence on turn efficiency as measured by tap-to-tap time. However, it 
 should be noted that the limited number of cases in our sample does not 
 enable us to reject the hypothesis that all differences among the obtained 
 correlations are a matter of chance. 
 
 The fact that mean turn tap-to-tap time is negatively related to the 
 mean ASo of the turn's four foremen suggests the possibility that ASo 
 scores within turns may be homogeneous. A ranked analysis of variance 
 test shows this to be the case. The long-lived nature of these groups could 
 cause this co-variation of ASo, either as a result of selective factors in 
 the personnel placement process, or because changes in interpersonal 
 perception occur as a result of group processes within the turns. The 
 possibility of such changes is suggested by the recent findings of Steiner 
 and Dodge (7) which indicate that Assumed Similarity may be in part a 
 function of the nature of the group. 
 
 While the present data support Hypothesis I rather than Hypothesis II, 
 a current study of another non-military group, viz., an analysis of board- 
 management relations cf a farm cooperative (4), gives results consistent 
 with Hypothesis Ila (which, as pointed out above, may be regarded as a 
 special case of Hypothesis I). However, the farm cooperative study also 
 does not provide any evidence in support of Hypothesis lib. Barring the 
 possibility that previous findings were due to chance, this suggests that 
 Hypothesis lib may be particularly germane to the military situation, or 
 to some special condition inherent in it, such as the short-lived nature of 
 the groups, or the restricted freedom of choice in "who works with whom." 
 It therefore seems advisable to use military groups for further validation 
 of Hypothesis II. 
 
 Additional Findings 
 
 In addition to the main criterion of tap-to-tap time, we obtained per- 
 formance ratings by shop superintendents and objective indices related to 
 the quality of output. The reliability of these criteria, their interrelations, 
 and their relation to the interpersonal perception score, ASo, are briefly 
 summarized in this section. 
 
• 
 
 
 
 
 
8 
 
 Ratings . Each shop superintendent and assistant shop superintendent 
 ranked all the foremen and melters in his shop on safety-mindcdness; 
 overall efficiency or productivity; and ability to handle men and maintain 
 high morale. The men were ranked by position, i.e., each superintendent 
 rank-ordered the four Senior Melters in his shop, then rank-ordered the 
 four General Foremen, and so forth. Reliability was estimated by com- 
 paring the Superintendent's ratings of Junior Melters with those of his 
 assistant superintendent. Table 2 shows these correlations for each of 
 the three rankings on safety-mindedness, productivity, and personnel 
 management skills. These estimates indicate generally high rater agree- 
 ment on the latter two criteria, but rather low rater agreement on the 
 safety-mindedness criterion. 
 
 TABLE 2 
 
 Correlations (Rho) between Superintendents' and Assistant 
 Superintendents' Ratings of Junior Melters 
 
 Criterion Correlation by shop (rho) Median 
 
 (rho) 
 
 A B C D 
 
 (N=8) (N=8) (N=8) (N=4) 
 
 Productivity 
 
 
 
 •95., 
 
 
 .90 
 
 ,92 
 
 
 .20 
 
 
 .91 
 
 Safety mindedness 
 
 
 
 .21 
 
 
 -57 
 
 .50 
 
 
 .20 
 
 
 .36 
 
 Personnel man- 
 
 
 
 
 
 
 
 
 
 
 
 agement 
 
 
 
 .79 
 
 
 .93 
 
 .95 
 
 
 .40 
 
 
 .86 
 
 Objective recor 
 
 ds 
 
 on 
 
 quality 
 
 of 
 
 output. 
 
 In additi 
 
 on 
 
 to the 
 
 output 
 
 critei 
 
 tap-to-tap time, the company maintains records on two secondary produc- 
 tion criteria which are essentially indices of the quality of the heat. The 
 first of these indices, called skulls , is the proportion of heats which, be- 
 cause of inadequate temperature control, leave a residue of steel in the 
 

ladle. The second score, called distressed heats , is the proportion of 
 heats which fail to meet the prescribed chemical specifications of the in- 
 tended consumer. Distressed heats are not total waste since they are 
 usually marketed to other consumers with less exacting chemical require 
 rnents or remelted for scrap. It should again be emphasized that skulls 
 and distressed heats are considered much less important than the tap-to- 
 tap criterion by company officials. 
 
 As can be seen in Table 3, the reliabilities of these criterion scores 
 are fairly high and compare favorably with tap-to-tap time. 
 
 TABLE 3 
 
 Reliability Estimates of Junior Melters 1 Tap Time, Skulls, 
 and Distressed Heat Criteria 
 
 Criterion Reliability estimates 
 
 Junior melters Turn average 
 
 N = 28 N = 16 
 
 Tap-to-tap time .82 .82 
 
 Skulls .84 .82 
 
 Distressed heats .54 .72 
 
 The intercorrelations among the three objective criterion indices and 
 productivity ratings are presented in Table 4. It will be noted that the 
 superintendents' ratings of productivity are not significantly related to any 
 objective production indices. This is all the more remarkable since the 
 superintendents' offices are responsible for collecting these productivity 
 data ard for forwarding them to the main office. These findings raise the 
 question anew whether we should place any credence in productivity ratings 
 by superiors, and to what foreman behavior or attributes these ratings 
 refer. 
 
10 
 
 TABLE 4 
 
 Intercorrelations of Junior Melters' and Turn Tap Time, Skulls, 
 Distressed Heat Criteria and Productivity Ratings** 
 
 Skulls Distressed heats Productivity- 
 ratings 
 
 Over Junior Melters (N=28) 
 
 Tap -to -tap time -.46* -.16 .14 
 
 Skulls .17 -.16 
 
 Distressed heats .12 
 
 Over turns (Nfl_6) 
 
 Tap-to-tap time -.26 .00 -.41 
 
 Skulls -.53* .10 
 
 Distressed heats .13 
 
 *P < .05 
 
 All criterion scores have been converted to T scores and where 
 necessary reversed so that high scores indicate good performance. These 
 intercorrelations are based on analyses of over 25,000 heats. 
 
 It is also worth pointing out that tap-to-tap time and heat control 
 (skulls) criteria computed for Junior Melters are negatively related, while 
 the heat control and metallurgical quality (distressed heats) criteria are 
 negatively related over turns (see Table 4). While we cannot be certain 
 of all the implications of these findings, this, as well as other studies which 
 utilize objective criterion scores (3, 5), suggest that team performance 
 may not be general over different tasks or even over different aspects of 
 the same task. A group which performs well on one criterion may or may 
 not perform well on another. This means that generalizations cannot be 
 drawn about "effective team performance" in general, but only about ef- 
 fective team performance on a given task X as measured by specific 
 
11 
 
 operations a, b, c, ..., etc. This is a severe limitation to development of 
 a general theory of group effectiveness but may be of only minor impor- 
 tance in a given field setting such as this, in which the prediction of a 
 single criterion appears justified because of economic considerations. 
 
 Table 5 shows the correlations between ASo of the various supervisors 
 and skulls, distressed heats, and productivity ratings. Only one of these 
 15 correlations reaches an acceptable significance level, viz., between 
 production ratings and ASo of Stock Foreman. The correlation indicates 
 that Stock Foremen with high ASo (small perceived differences between 
 most and least preferred co-workers) tend to be rated good on productivity. 
 Since only one correlation of 15 reaches the five per cent significance 
 level, no importance can be attached to this finding without further inves- 
 tigation. 
 
 TABLE 5 
 
 Correlations (Rho) between ASo of Various Supervisors and Skulls, 
 Distressed Heats and Productivity Rating Criteria 
 
 Supervisor 
 
 Skulls 
 
 Distressed 
 
 Productivity 
 
 ASo** 
 
 
 heats 
 
 ratings 
 
 General Foreman 
 
 -.38 
 
 .38 
 
 .03 
 
 Stock Foreman 
 
 -.01 
 
 .09 
 
 .62* 
 
 Fit Foreman 
 
 .30 
 
 -.02 
 
 -.40 
 
 Senior Melter 
 
 .37 
 
 .00 
 
 -.20 
 
 Supervisors' 
 
 -.06 
 
 .13 
 
 ,28 
 
 average 
 
 
 
 
 *p < .05 
 **N varies from 13 to 16, due to missing data. 
 
12 
 
 S ummary and Conclusions 
 
 A study was conducted relating the interpersonal perceptions of open 
 hearth shop foremen to the productivity of their work units. Interpersonal 
 perception was measured by means of Assumed Similarity (ASo) tests 
 which reflect how similar or different a person describes his most and 
 his least preferred work-companions. Group effectiveness measures were 
 based on output as indicated by "tap-to-tap" time, the time required to 
 complete a "heat" of steel. This criterion measure has considerable sta- 
 bility and is regarded as the most important production index by company 
 officials. 
 
 Management personnel of four open-hearth shops of a large steel com- 
 pany participated in this study. Interpersonal perception (ASo) and socio- 
 metric tests were administered to all available Ss. 
 
 Significant relations were found between supervisor ASo and the tap- 
 to-tap time index. These results are consistent with the hypothesis that 
 more effective groups have supervisors who perceive relatively large 
 differences between their most- and least-preferred co-workers. The 
 study did not support the alternative, partially overlapping hypothesis that 
 more effective groups have supervisors who maintain a moderate amount 
 of psychological distance between themselves and their co-workers. 
 
 Additional criteria of secondary importance, including ratings by 
 superintendents and criteria related to the quality of the output, were 
 obtained and analyzed. The reliability, interrelations, and the relation- 
 ships of these secondary criteria to ASo are discussed in light of the 
 primary data of this study. 
 

 
13 
 References 
 
 1. Qronbach, L. J., and Gleser, Goldine C. Assessing similarity between^ 
 
 profiles. Psychol . Bull ., 1953, 50, 456-473. 
 
 2. Fiedler, F. E. Assumed similarity measures as predictors of team 
 
 effectiveness, j". abnorm . soc . Psychol ., 1954, 49, 381-387. 
 
 3. Fiedler, F. E. The influence of leader-keyman relations on combat 
 
 crew effectiveness. _J. abnorm . soc . Psychol ., 1955, fU, 227-235. 
 
 4. Godfrey, Eleanor P., Wrigley, C. F., Mannheim, Betty, Hall, D. M., and 
 
 Fiedler, F. E. The effect of interpersonal relations on the success 
 of consumer cooperatives. Paper read at the American Sociolo- 
 gical Society, Washington, D. C., September, 1955. Research 
 supported by Contract N6-ori-07l35 between the Univer. of Illinois 
 and the Office of Naval Research. 
 
 5. Knoell, Dorothy, and Forgays, D. G. Interrelationships of combat crew 
 
 performance in the B-29. USAF Hum . Resour. Res . Cent ., Res . 
 Note , 1952, CCT 52-1. 
 
 6. Osgood, C. E., and Suci, G. A measure of relation determined by both 
 
 mean difference and profile information. Psychol . Bull ., 1952, 
 49, 251-262. 
 
 7. Steiner, I, D., and Dodge, JoanS. Interpersonal perception and role 
 
 structure as determinants of jgroup and individual efficiency. 
 Hum. Relat., in press. 
 
m 
 
 s 
 
 i 
 
 - 
 
 i 
 
 H 
 
 **