Sex Roles. VoL 2, No. 1, 1976
Achievement Motivation in College Women Revisited: Implications for Women, Men, and the Gathering of Coconuts z Lynette K o h n Friedrich2
Pennsylvania State University
An attempt was made to replicate the findings o f French and Lesser (1964) within one university on the effects o f value-appropriate arousal conditions on achievement motivation scores and motivation performance relationships in women. The findings o f this study do not uphold those previously reported. There were no significant differences on either achievement motivation scores or task performance under value-appropriate versus -inappropriate arousal conditions. In the present study there were two innovations --performance criteria were extended to include academic performance, and scoring procedures were developed to give equal weight to achievement imagery in areas defined as traditionally feminine. The principle findings concern the importance o f the scoring procedures. Intellectual (classic) N Ach does not relate to value orientation, task performance, or grade point average. Women's Role N Ach, however, is significantly related to task performance and to grade point average. This motivation-performance relationship exists despite differences in value orientation among women. It is suggested that when scoring procedures allow subjects, male or female, to express achievement strivings in areas o f value to them, then performance may be morc sensitive to motivational variables. JThis article is based on research supported by Grant 1 ROI HD-02403-01 from the National Institute of Mental Health to Lynette K. Friedrich and John S. Harding and by funds from the College Committee on Research, New York State College of Human Ecology, Comell University, to Lynette K. Friedrich. The author is deeply indebted to Dr. John Harding for faculty support in securing the ~ a n t and for his invaluable participation as consultant. Dr. Doris B. Rosen and Dr. Elizabeth Anisfeld made significant contributions to the study. In addition, the assistance of Serena Weaver, Helen Grebow. Bruce Ambler, and Jansis Smithels is gratefully acknowledged. Further acknowledgment is made to Dr. Elizabeth G. French for her advice and generous cooperation. 2 Requests for reprints should be sent to Lynette Kohn Friedrich, Department of Psychology, The Pennsylvania State University, University Park, Pennsylvania 16802. 47 © 1 9 7 6 P l e n u m PuDIIshing C o r D o r a t i o n , 2 2 7 West 1 7 t h Street, N e w Y o r k , N . Y . 1 0 0 1 1 . N o I~art o f t h i s pul01ication may Oe reproduceo, stored in a r e t r i e v a l system, or t r a n s m i t t e c l , in any f o r m or b y any means, electronic, mechanical, photocopying, microfilming, recording. or o t h e r w i s e , w i t h o u t w r i t t e n permission o f t h e Dublisher.
48
Ftiedrich
The controversy over the meaning and validity of measures of achievement motivation in women has grown in complexity over the past 25 years. Thelma Alper (1974) recently reviewed the "now-you-see-it-now-you-don't phenomenon" in female achievement motivation research and raised questions concerning the effects of differences in methods. Among the variables identified which may account for inconsistent findings are sex-role orientation and scoring procedures. The data for the present study were coUected as a part of a research project on women's achievement conducted in 1968 (Friedrich & Harding, 1968). The effects of value orientation on the achievement motivation-performance relationship and scoring procedures which include feminine goals in achievement imagery were investigated. One major goal was to test some of the findings of French and Lesser (1964). French and Lesser, arguing that standard arousal speeches were more appropriate to the value structures of males than females, included in their theoretical formulation the effect of value orientation on motivation level and performance. They found that women responded with elevated achievement motivation scores only when they heard an arousal speech that corresponded to a value orientation in which they were high. Also, the level of output on a performance test which followed the projective test was positively correlated to the achievement motivation scores only when the arousal condition was value appropriate, but not otherwise. They used homogeneous samples at six colleges to represent different value orientations. In order to test these findings, some of the features of measurement and design used by French and Lesser were retained. However, in the present study value orientations were assessed within one university- Comell. In addition, performance criteria were extended to include academic performance, and scoring procedures were developed to give equal weight to achievement imagery in areas defined as traditionally feminine. Therefore, it was possible to investigate broader relationships: value orientation, and two categories of achievement motivation to academic performance, as well as task performance.
METHOD Three types of instruments were employed: (1) a measure of value orientation, (2) a measure of achievement motivation, and (3) a measure of output on an achievement-related performance test.
Measurement of Value Orientation The Student Attitude Scale, developed by French and Lesser, was a 65item true-false questionnaire with two subscales to measure the extent to which
Achievement Motivation in College Women
49
a subject values various aspects of intellectual attainment, e.g.: Most girls' plans for after graduation include a career in which they can develop and use theix talents.
and the extent to which she values various aspects of woman's role, e.g.: The gi~ls feel that an important function of college should be preparation for the social responsibilities most o f them will assume when they marry.
The items were designed to avoid the necessity of a choice between the two value orientations, and it is theoretically possible to obtain maximum scores on both scales without being inconsistent. 3 The Student Attitude Scale was revised in two ways for the Cornell sample: The "most girls here" format was changed to the "10 girls you know best at Cornell." The "10 girls" structure allowed for differences in value orientations in one university and was chosen to avoid defensiveness in self or friends' responses. On the basis of pretesting, the assumption that the subjects answered in terms of their own values was given substantial support. Item analysis of Woman's Role scale and Intellectual scale responses for 125 pretest subjects was done to select items which discriminated well on the two scales for high and low scoring groups. Five items were deleted and ten new items added to the Woman's Role scale to parallel the discriminating items. The final test contained 69 i t e m s - - 3 0 keyed on the Intellectual scale; 39 on the Woman's Role scale.
Measurement of Achievement Motivation The Test of Insight (French, 1958) - Form 1II with female names -- was employed to measure the level of motivation. This test consists of 10 single-sentence descriptions of behavior purported to be characteristic of a hypothetical individual. The test is described as a measure of insight into the behavior of others and the subjects are asked to "explain" the behavior described. The number of achievement-relevant ideas expressed in the explanation makes up the scores. The test items are: l) Sue frequently organizes groups or committees. 2) Judy said "Look what I've done!" 3) Carol Nves lots of parties. 4) Clara always lets the "other fellow" win. 5) Mary spends lots of time just socializing. 6) Mice is always trying something new. 7) Sarah chooses her friends very carefully. 8)
~French and Lesser employed the "most girls here" format and, following the Student Attitude Scale, gave a series of questions which measured the extent to which women felt they were like the "'girls here." Subjects with low identification scores were dropped. Six colleges were then selected on the basis o f homogeneity o f scale scores -- two colleges for each value orientation group: High Intellect-High Woman's Role, High I n t e l l e c t - L o w Woman's Role. and Low Intellect-High Woman's Role.
50
Friedrich
Dorothy never joins clubs or social groups. 9) Barbara is always willing to listen. 1O) Ruth said, "They probably won't ask me to go with them."*
Scoring System In the Cornel] study, scoring procedures were developed to give equal weight to traditional achievement motivation imagery and to a new category of feminine achievement motivation imagery. The Intellectual-Achievement score (referred to in this paper as I N Ach) was obtained by use of the scoring method originally devised by French for the Test of Insight (French, 1958). This method involves scoring categories which overlap to a high degree with those of McClelland, Atkinson, Clark, and Lowell (1953), but are better suited to the short-sentence stimulus material of the Test of Insight. I N Ach scores, then, are very similar to the classic N Achievement scores. Woman's Role N Achievement scores (referred to as WR N Ach) were based on the same system of categories used to score I N Ach imagery, but the content dealt with traditional feminine role activities. Imagery which centered about marriage, heterosexual popularity, social leadership, excellence in feminine skills, civic leadership, etc., was coded as WR N Ach imagery when it met the category criteria. Care was taken to avoid scoring imagery which was affiliatire rather than achievement oriented. References to "valuing friendship" or "enjoying people" were not scored as achievement imagery; there had to be clear evidence of achievement aspects, e.g., "She tried to make each party a unique SUCCESS.''s
Since the French method of scoring protocols has not been published, scoring reliability for I N Ach was first established on the McClelland, et al. method (Atkinson, 1958). The correlation coefficient for the reliability among the five raters was rs = .97. Rules for the French scoring method were developed with the help of Elizabeth French. To achieve a comprehensive scoring system which encompassed WR n Ach as well as 1 N Ach, however, it was felt that the whole range of responses elicited in our test papers should be scanned. All of the tests were read (without identi4The Test o f Insight was used in the French and Lesser study, although male and female names were employed on two different forms and under neutral conditions and arousal conditions. Cornell comparisons are made only with their data from the female form o f the test, under two arousal conditions. SThe concept of WR N Ach was introduced by Elizabeth French. French and Lesser (1964) tried experimental, scoring o f the new category. Their data are presented in terms o f totals rather than two separate achievement imagery scores. Therefore, it was necessary to develop rigorous scoring procedures for the Comell study. A guide for the scoring of I N Ach and WR N Ach may be obtained from the author.
Achievement Molivalionin Coileg~Wome~
$1
fying data) and two sets of rules were finally developed - general rules and specific rules for each o f the 10 questions on the Test o f Insight. Reliability coefficients were computed to determine interrater reliability by the Analysis of Variance formula. In the final practice session an overall reliability coefficient o f .93 for the five judges on 17 protocols was obtained. The procedure for the actual scoring o f the test protocols was to assign the same two questions to each o f the five judges. In this way it was hoped to randomize idiosyncratic scoring biases by judges, since the total score on each protocol was based on five separate judgments. All questions were then scored again by a second judge. Since the above procedure was used, the interrater reliability coefficients were also c o m p u t e d for the final practice session on each question on all protocols. These coefficients ranged from rs = .86 to rs = .98.
Measurement of Performance on Tests Two performance test were used to measure achievement-related effort. The Scrambled Words Test was chosen as relevant to Intellectual goals and followed the Intellectual arousal statement. The Social Problems Test was designed to relate to Woman's Role goals and followed the Woman's Role arousal statement. The Scrambled Words Test is an anagrams test constructed after Lowell (1952) from the first three levels o f the Thorndike-Lorge word list. The test is considered a measure o f effort rather than ability for college students due to the familiarity o f the words. There were 10 pages containing 20 words each. Subjects were allowed 2 minutes per page and the time limits were emphasized in the instructions, The test was scored by allowing 1 point for each 2 words solved .6 The Social Problems Test is also designed to be a measure of effort. Ten items describing social situations in which the subject might find herself were given and the subject was to list appropriate courses o f action she could take (e.g., "Your husband is unhappy with his present j o b but hesitates to leave because he feels concerned about supporting his family. What can y o u do to help him?"). Subjects were allowed 2 minutes per item and the time limits were again emphasized in the instructions. 7 6The Scrambled Words Test employed in the CorneU study was identical to the one used by French and Lesser. ~While the structure and design of the French and Lesser Social Skills Test were preserved in the Cornell Social Problems Test, items were revised in order to present unstereotyped social situations which would generate effort rather than hostility in a highly selected sample. Minor alterations in wording were made for several items, but for others the item was dropped (e.g.. "You are a bride and you want to help your husband get ahead as much as you can. What things do you think about -- both long-term and immediate?"). A new item was written on a similar theme ~see example above).
52
Friedrich COLLECTION OF DATA
Sub/ects All 1,459 women from the sophomore, junior, and senior classes at CorneU University were asked to take the Student Attitude Scale on a voluntary basis. The Student Attitude Scale was given to 840 subjects and scores were obtained on each of the two subscales (Intellectual value orientation and Woman's Role value orientation). Frequency distributions were made for each o f the subscales separately, and each distribution was divided approximately into thirds: "high," " m e d i u m , " and "low." A bivariate frequency distribution was then made by plotting each subject's score on the Woman's Role scale against her score on the Intellectual scale to see whether cutting points thus established would yield enough subjects in the two-way classification groups that were to be used in the remainder o f the study - i.e., High I - H i g h WR, High l - L o w WR, Low I - H i g h WR. The actual cutting points used to establish the High value orientation groups were 19 and above on the Woman's Role scale and 26 and above on the Intellectual scale. The cutting points used to establish the Low value orientation groups were 13 and below on the Woman's Role scale and 22 and below on the Intellectual scale, s Of the 450 subjects whose scores fell into the desired combination o f value orientations, 325 returned for a second session. In the latter session, half of each group was given a motivation test and performance test under intellectual arousal and half under Woman's Role arousal conditions. Subjects were randomly assigned to arousal treatments and means tested to assure that the group means on Woman's Role and Intellectual scores were not significantly different.
Administration of the Student A tt~tude Scale In order to achieve standardized testing conditions, the Student Attitude scale was given on one evening to the women in a specified room in their residences. The testing sessions were supervised, although all instructions were printed on the cover sheet so that no verbal instructions would be required.
*In terms of comparability with the French and Lesser scores, the Cornell scores on Woman's Role value orientation are quite similar. The Intellectual value scores are similar, although the skewed Intellectual value scores present a less sharp contrast for the High l-Low WR group than that of French and Lesser. Cornell selection criteria as well as the use of signed answer sheets are seen as contributing factors to the skewed Intellectual value scores.
A c h i e v e m e n t Motivation in College W o m e n
53
Reliability analyses conducted on the results of the revised Student Atti. rude scale for the 840 subjects indicated that the reliability coefficients for both scales were considerably elevated over pretest coefficients. Subscale reliabilities of .59 (Woman's Role scale) and .66 (Intellectual scale) had been obtained in the pretest. The revised questionnaire subscale reliabilities were .80 for the Woman's Role scale and .81 for the Intellectual scale. A Pearson r of - . 3 6 between the two subscales indicates a substantial negative relationship, but they are relatively independent of each other.
Experimental Treatment The subjects from the desired value orientation groups were contacted individually and asked to come to a classroom on one of two evenings. The subjects were not aware of the selection criteria. They were told that the procedure would take about an hour and that they would be paid for their time. The author was the experimenter for all sessions. Intellectual Arousal Condition. The experimenter made the following statement: This is the second part of the s t u d y y o u took part m a few weeks ago. There are two tests we would like you take tonight. The first is called the Test o f Insight. T h e main test is called the Scrambled Words Test. it measures one aspect o f vocabulary -- facility with words. As you m a y k n o w , vocabulary has been found to be the best single index of general intelligence we have. So you can see that it is very i m p o r t a n t for you to do as well as y o u can o n this test. It has been given to girls at a n u m b e r o f colleges all over the c o u n t i y so we can make comparisons of general inteUigenco level at these colleges. Naturally, we expect the girls at Cornell to do very well. So please try your best.
Subjects were then given the Test of Insight. When they had fimshed, part of the arousal speech was repeated and the Scrambled Words Test administered? Woman's Role Arousal Condition. The experimenter made the following statement: This is the second part o f the s t u d y y o u took part in a few weeks ago. There are two tests we would like you to take tonight. The first is called the Test o f Insight. T h e main test is called the Social Problems Test. To most girls a successful marriage, children, and a full and happy family and social life is a very important goal. To attain this, a girl n e e d s to develop certain skills to enable her to handle herself in various social situations, make friends and eventually to manage a home and a family. T h e test y o u are going to take, the Social Problems Test, is a measure o f resourcefulness in handling social problems and tends to predict how successful a girl will be in foreseeing and dealing with these problems.
*The a~ousal s t a t e m e n t and procedures are identical to those used by French and Lesser.
54
Friedrich
The Test of Insight was then given, followed by a repetition o f part of the arousal statement and the Social Problems TestJ o
RESULTS Arousal Conditions and Achievement lmagery The first hypothesis from French and Lesser asserts that achievement motivation will be higher when arousal cues are related to a goal which is value relevant for the subjects than when arousal cues are related to a nonrelevant goal. In terms of the groups obtained from the Student Attitude scale, the prediction is that High Intellect-Low Woman's Role value orientation group subjects should have higher achievement motivation scores under Intellectual Arousal than under Woman's Role Arousal. Subjects in the Low Intellect-High Woman's Role value orientation group should have higher scores under Woman's Role Arousal than under Intellectual Arousal. H The hypothesis was tested separately by a 2 X 2 analysis o f variance for each type o f achievement motivation -- I N Ach and WR N Ach. Table I presents the mean motivation scores on I N Ach for the two contrasting value orientation groups under Appropriate and Inappropriate Arousal conditions. The analysis of variance indicates that there are no significant differences for the value orientation groups (F = .06), appropriateness of arousal (F = .42), or for the interaction of the two variables (F -- .00). In Table II the mean scores on WR N Ach for the two value orientation groups under Appropriate and Inappropriate A~'ousal conditions are given. While nonsignificant, the effect of Arousal condition on Women's Role achievement motivation is in the predicted direction (F = .74). There are higher N R N Ach scores for both groups under Appropriate Arousal conditions than under Inappropriate Arousal conditions. Thus, the results for both achievementmotivation measures - I N Ach and WR N Ach -- are in the same direction as
~°$tatements made by the experimenter differ very slightly from those used by French and Lesser in the presentation of the Social Problems Test. T h e test is presented as a measure of " r e s o u r c e f u l n e s s " in handling social problems rather than " a t t a i n m e n t " o f social skills. All other features o f the original instructions are preserved. '~ It should be noted that the two groups High l - H i g h WR and High l-Meal. WR were not used to test Hypothesis I. It was not originally postulated that these value orientation groups would differ in their response to t r e a t m e n t conditions.
Achievement Motivation in College Women
55
Table I. Summary of Means for Intellectual N Ach Scores by Value-Orientation Group and Appropriateness of Arousal Treatment Arousal treatment Value orientation group
Appropriate arousal
High I - L o w WR Low l-High WR
54 46
Ns
Average
Inappropriate ~ousal
Average
Ns 6.72 6.63
59 40
6.68
6.42 6.23
6.57 6.43
6.32
6.50
those reported by French and Lesser, but they are so small as to be statistically nonsignificant.
Arousal Conditions and the Motivation Performance Test Relationship The second hypothesis from French and Lesser states that achievement motivation scores will be significantly related to measures of performance only if the achievement motivation scores are obtained under appropriate arousal conditions and the performance tests are value appropriate. The intercorrelations of motivation (total N Ach, I N Ach, and WR N Ach) and performance measures (Scrambled Words Test - SW -- and Social Problems Test -- SP) by arousal condition appear in Table I11. Tests of significance indicate no significant differences between Appropriate and Inappropriate Arousal treatment groups in the intercorrelations of per-
Table ii. Summary of Means for Woman's Role N Ach Scores by Valu©-Orientation Group and Appropriateness of Arousal Treatment Arousal ~eatment Value
orientation group
High l-Low W R Low l-High WR
Average
Appropriate arousal
Inappropriate arousal
Ns
Ns
54 46
6.44 8.02
7.23
59 40
Average
6.05 7.55
6.24 7.78
6.80
7.01
56
Fnedrich Table !1I. lntercorrelations of Performance and Motivation
Measures in Groups Receiving Appropriate and Inappropriate Arousal Conditions I
Ns
SW
SP
N Ach
Wg N Ach
Groups receiving appropriate arousal conditions 149 80 68 149 149
1. 2. 3. 4. 5.
Total N Ach SW SP 1N Ach WRN Ach
-.12
.22
.78 -. 14 .14
.80 -.06 .19 .25
Groups receiving inappropriate arousal conditions 99 40 59 99 99
I. 2. 3. 4. 5.
Tot~NAch SW SP I NAch WRNAch
-.18
.33
.83 -.08 .20
.75 -.18 .38 .26
formance and achievement motivation measures; and the direction of the differences is usually counter to the hypothesis, that is, higher correlations under Inappropriate rather than Appropriate arousal conditions. In summary, the Cornelt findings do not uphold those of French and Lesser on the effect of value-appropriate arousal on either achievement motivation or the achievement motivation-performance relationship.
Innovations of Cornell Study A second purpose of the study was to investigate the relationship of (a) achievement.motivation scores (both I N Ach and WR N Ach) and (b) value orientation to academic performance. It is in the consideration of these relationships that the effect of scoring of achievement imagery becomes apparent. The intercorrelations of the nine major variables used in the Cornen study appear in Table IV. The variables include Woman's Role value-orientation score (WR score), Intellectual value-orientation score (I score), verbal Scholastic Aptitude Test scores (SAT), grade point average (GPA), total N Ach, WR N Ach, I N Ach, Scrambled Words Test score (SW), and Social Problems Test score (SP).
Achievement Motivation and Performance In Table IV there is a rather striking contrast between I N Ach and WR N Ach. I N Ach does not relate significantly to any o f the other variables. WR N
Achievement Motivation in College Women
It
57
:
I
I
t~
II
It
I
°_
z I
?
m
tt
r..
v
v
58
Fried_rich T a b l e V. Partial Correlation between Achievement Motivation (I N Ach then WR N Ach) and Achievement (Grade Point Average) with. Ability (SAT Scores) Held Constant
Variable 1 - I N Ach Variable 2 - GPA Variable 3 - SAT
Variable 1 - WR N Ach Variable 2 - GPA Variable 3 - SAT
rt2.3 = - . 0 2
rt:.3 = .14
N = 312 n.s.
N= 312 p = < .05
Ach, by contrast, is significantly related to grade point average (.12 p < .05) and to the appropriate performance t e s t - t h e Social Problems Test (.26 p < ,01). The distinction between I N Ach and W R N Ach was further clarified when the hypothesis was tested that there is a relationship between achievement-motivation scores and academic ackievement when ability is held constant. The results of the partial correlation are presented in Table V. t 2 While the relationship between WR N Ach and academic performance is modest, it is consistent with the positive relationship found between WR N Ach and another measure of performance - the Social Problems Test. I N Ach does not relate to any o f the performance measures. The latter finding supports previous findings for females employing traditional achievement-motivation scoring procedures (Cole, Jacobs, Zubok, Fagot, & Hunter, 1962; Sundheim, 1962).
Value Orfentation and Academic Performance From Table IV it can be seen that Woman's Role value orientation is negatively related to SAT scores and grade point averages at the .01 level of significance. Intellectual value orientation is mildly related to SAT scores (.08 p < .05) but is not related to the other variables. When the hypothesis was tested that value orientations will be related to academic achievement with ability held constant, a similar finding emerged. Partial correlations including value orientation, grade point average, and SAT scores for all 77t subjects indicate that Woman's Rote value orientation tended to be negatively related to academic achievement (rt2.a = - . 0 7 ; p < .05). Intellectual value orientation was unrelated. ~3 ,2 The scores of all subjects were used to test this Hypothesis: no restrictions were imposed by value-orientation group or by appropriateness of arousal treatment. '~When the scores from the highly selective a~ts college sample are anaJyzed separately (N = 395), thc contrast in the ~elationship of value orientation and academic performance is accentuated. Intellectual value orientation correlates .05 with grade point average when SAT is held constant, while Woman's Role value odentalion correlates - . I 3 with GPA under the same restrictions. This latter correlation differs from zero at the .01 level of significance.
Achievement Motivation in College Women
59
WR N Ach, Woman's Role Value Orientation, and Performance Although Woman's Role value orientation and WR N Ach are positively related (Table IV), a clear distinction must be made between them. Woman's Role value orientation tends to be incompatible with academic achievement, while WR N Ach tends to relate positively to academic achievement and the Social Problems Test. Consideration of performance on the Social Problems Test helps to clarify the value orientation-motivation-performance relationship further. There is a positive correlation between the Social Problems Test and WR N Ach (.26 p < .01). When this relationship is viewed by separate value orientation groups (Table III) it is higher for those women with low Woman's Role value orientation scores (.38 Inappropriate) than for women with high Woman's Role value orientation scores (.19 Appropriate). The findings suggest, then, that value orientation does have an impact on performance. But the more important implication is that a motivation-performance relationship exists despite differing value orientations for women when the feminine goals and social skills expressed in achievement fantasy are scored. In summary, the Comell findings are in accord with those of many other researchers who have not found I N Ach, or traditional N Ach, to be a useful or meaningful measure with women. Nor do the findings uphold the interpretation of French and Lesser that value-appropriate arousal speeches reconcile the problem. The important implication from this study is that WR N Ach appears to be a more appropriate measure of achievement motivation fantasy for all of these women - despite their value orientation. It relates positively with a performance task and with grade point average with ability held constant and presents a consistent, if modest, profile.
DISCUSSION
Implications for Methods in Study of A chievemen t Mo tivation The CorneU findings suggest the need for a broader view of the measurement of achievement motivation. The traditional scoring of achievement-related fantasy confined the concept of competition with a standard of excellence to culturally approved male pursuits and modes. Females, in the Cornell study, were more likely to express achievement motivation in areas that are defined traditionally as feminine, particularly social skills. The attainment of excellence in social skills or interpersonal relations may also be part of an internal standard of excellence, although not one which has been accepted in the middle class
60
Friedrich
entrepreneurial def'mition which has dominated the research on achievement motivation. A question must, then, be raised about the fundamental conceptualization of achievement motivation. If the argument is that people express achievement motivation fantasy in the areas which are accessible to them and which they value, the past failure of achievement-motivation literature to account for women is not surprising. However, unless achievement motivation is conceived of as a magic quantity existing in a timeless void, the achievement stnvings of both women and men in areas in which they are personally involved must be considered. What is personally involving for the two sexes has changed in the past and is changing. Although classic achievement-motivation scoring procedures appear to have restricted imagery to the male mode, there was an awareness of possible timebound, middle-class bias. A cautionary note is sounded in the scoring guide (Atkinson, 1958, p. 721): "AI. Long-term achievement interest: 'he has wanted to become a garage mechanic.' Remember in scoring not to impose your own sense of achievement values on the characters. In stories written by some primitive island people, there might be long-term interest of an achievement-related sort in climbing trees for coconuts." While the matter of achievement imagery in primitive men is far from the major concern of this discussion, the warning is relevant. Beware of limiting the WR N Ach scoring category to females. There are numerous indications in the recent literature that men are rejecting the conventional goals which have been scored in achievement imagery or show achievement responses which inc]ude interpersonal skills (Feather & Simon, 1973; Lunnenborg & Rosenwood, 1972; Morgan & Mausner, 1973; Robbins & Robbins, 1973). While it is suggested that young males and females are rejecting middleclass, sex-stereotyped success conventions, much of the discussion tends to be buried in the post hoc sections of the literature. There is need for systematic investigations of both sexes which are sensitive to rejection of stereotyped cues and avoid the pitfalls of conventionally male-oriented scoring restrictions and gross thematic categorizations. If the proverbial baby is not to be thrown out with the bath, the delineation of the different modes and areas in which achievement strivings can be expressed is of critical importance. More diverse and refined value-orientation assessments, cues, and scoring procedures are needed if the measurement of achievement motivation is to be of value in the prediction of behavior.
REFERENCES Alper, T. G. Achievement motivation in college women: A now-you-sec-it-now-you-don't phenomenon. American Psycholo~st. 1974, 29, 194-203. Atkinson. J. W. Motives in fantasy, action, and society. Princeton: Van Nostrand, 1958.
Achievement Motivation in College Women
61
Cole, D., Jacobs, S., Zubok, B., Fagot, B., & Hunter, 1. The relation of achievement imagery scores to academic performance. Journal of Abnormal and Social Psychology, 1962,
65, 208-211. Feather. N. T., & Simon, J. G. Fear of success and causal attribution. Journal of Personality, 1973,41,525-542. French, E. G. Development of a measure of complex motivation. In J. W. Atkinson (Ed.), Motives in fantasy, action, and society. Princeton: Van Nostrand, 1958, pp. 242-248. French, E G., & Lesser, G. S. Some characteristics of the achievement motive in women. Journal of A bnormal and Social Psychology, ! 964, 68. 119-128. Friedrich, L. K., & Harding, J. Achievement motivation and academic performance in women. Final report. June 1968, Comell University, Grant No. I ROI HD-02403-01, National Institute of Mental Health. Lowell. E. L. The effect of need for achievement on learning and speed of performance. Journal of Psychology, 1952.33, 31-40. Lunnenborg. P. W., & Rosenwood, L. M. Need affiliation and achievement: Declining sex differences. Psychological Reports, 1972, 3l, 795-798 McClelland. D. C.. Atkinson, J. W., Clark, R. A,, & Lowell, E. L. The achievement motive. New York: Appleton-Century-Crofts, 1953. Morgan, S. W., & Maumer, B. Behavioral and fantasied indicators of avoidance of success in men and women. Journal of Personality, 1973, 41,457-470. Robbins. L.. & Robbins, E. Comment on: "Toward an understanding of achievement-related conflicts in women." Journal of Social Issues, 1973, 29, 133-137. Sundheim, B. J. M. The relationship among "'n" Ach. "n'" affiliation, sex-role concepts, academic g~ades, and curricular choice. Unpublished doctoral dissertation, No. 63-1528. Columbia University, 1962.