University of North Florida UNF Digital Commons

All Volumes (2001-2008) The sprO ey Journal of Ideas and Inquiry

2001 Sex Differences in Mechanical Aptitude: An Investigation of Sex Differences in Mechanical Aptitude and Its Relation to Nonverbal Abilities Karolina E. Christensson Ringby University of North Florida

Follow this and additional works at: http://digitalcommons.unf.edu/ojii_volumes Part of the Social and Behavioral Sciences Commons

Suggested Citation Ringby, Karolina E. Christensson, "Sex Differences in Mechanical Aptitude: An Investigation of Sex Differences in Mechanical Aptitude and Its Relation to Nonverbal Abilities" (2001). All Volumes (2001-2008). 132. http://digitalcommons.unf.edu/ojii_volumes/132

This Article is brought to you for free and open access by the The sprO ey Journal of Ideas and Inquiry at UNF Digital Commons. It has been accepted for inclusion in All Volumes (2001-2008) by an authorized administrator of UNF Digital Commons. For more information, please contact Digital Projects. © 2001 All Rights Reserved Sex Differences in based on a smaller sample than originally planned. Mechanical Aptitude: An Little research has been conducted Investigation of Sex examining the basis for the reported sex differences in mechanical aptitude. Differences in Mechanical Mechanical aptitude tests are used Aptitude and Its Relation to extensively in selection procedures for Nonverbal Abilities mechanical . A study investigating the validity of two of these measures and Karolina E. Christens son Ringby the differential performance of each sex group on them is quite timely because Faculty Sponsor: Dr. Susana Urbina, more women are seeking in Professor of areas that require mechanical abilities.

Abstract Introduction

This study investigates sex differences The empirical literature suggests that in mechanical abilities. Numerous the construct of general , or the research findings support the existence of "," encompasses a wide spectrum of sex differences in mechanical aptitude as learned and innate human behavior and well as in other abilities. Recent research functioning made up of factors such as suggests that there is a relationship cognitive abilities, personal variables, and between mechanical aptitude and certain experience, as well as musical and nonverbal reasoning (e.g., visual-spatial) mechanical aptitudes (Anastasi and Urbina, abilities. Due to the possible links 1997). The g. or general, factor originally between these constructs, sex differences identified by Spearman, is placed by Vernon in a nonverbal measure of general at the top of the hierarchy of cognitive intellectual functioning, made up of five abilities followed by specific abilities such subtests, were also evaluated. Thus, the as mechanical , spatial ability, purpose of this study is twofold: to verbal ability and quantitative ability investigate sex differences in mechanical (Vernon, 1950). The g factor has proved to aptitude and nonverbal abilities, and to be useful in predicting criteria of explore the construct validity of all the performance as well as performance in measures employed. academic settings. Females and males were compared in Although the g factor is very useful, terms of their performance on two there are a number of situations in which it mechanical aptitude tests and one test of is desirable to learn about an individual's nonverbal abilities. It was expected that, profile of strengths and weaknesses in on average, men would do better on both intellectual abilities, for example, in of the mechanical aptitude tests. The vocational counseling. Therefore, we must construct validity of all the measures was also look at performance levels in abilities investigated by correlating the scores on that are narrower than g. such as mechanical the various instruments and by aptitude. According to Bennett (1994), comparing the results obtained with the mechanical aptitude "may be regarded as expectations suggested by the empirical one aspect of intelligence, if intelligence is literature on sex differences. Thus far, all broadly defined" (p. 8). However, tests the major hypotheses have been developed to assess general intelligence confirmed. However, these results are typically do not provide information on a person's standing in mechanical aptitude.

Osprey Journal of Ideas and Inquiry 53 There is a need for specialized instruments, BMCT assesses the constructs of such as mechanical aptitude tests, that mechanical information, spatial measure specific clusters of abilities visualization, and mechanical reasoning pertaining to certain tasks or occupations. (Tippins, 1991). Although mechanical aptitude is related to Recently, a new test of mechanical general intelligence, it also comprises a aptitude, the Wiesen Test of Mechanical complex yet relatively independent set of Aptitude (WTMA), was developed with the abilities. Wiesen (1999) suggests that intention of creating a fairer measure for mechanical aptitude is "a function of general women than the traditional measures used in intelligence, spatial ability, interest, and this area, such as the BMCT. Although mechanical knowledge" (p.3). Mechanical males still outperform females on the aptitude tests measure a set of complex WTMA, the male advantage is not as great cognitive abilities whose central components as on the BMCT (Chapman, 1998; Wiesen, are spatial abilities, mechanical reasoning 1999). and information, as well as perceptual In spite of these well-documented sex discrimination (Anastasi and Urbina, 1997). differences on mechanical aptitude tests, Related to individual performance little research has been conducted examining levels in cognitive abilities, and of equal the basis for sex differences on these importance, is the issue of individual and measures. Since more and more women are group differences in cognitive functioning. seeking employment within fields that Sex differences in cognitive abilities are one involve mechanical ability and since such difference of interest. Whereas there is mechanical aptitude tests are used no evidence of a sex difference in general extensively in selection for such jobs, it is intelligence, sex differences have been important to investigate and to better observed in broad cognitive abilities. For understand the reasons for sex differences in example, Halpern (1992) identifies verbal, various available measures of mechanical quantitative, and visual-spatial ability as the aptitude. The increasing number of women "main loci of sex differences" (p.64). Males seeking employment in mechanical excel in certain quantitative tasks and on occupations also underscores the need for most visual-spatial tasks. Females tend to valid and fair predictors of job performance have better verbal skills and outscore males in this field. on certain tests of quantitative abilities with The purpose of this study is twofold: to verbal components. In addition to the investigate the construct validity of all the documented sex differences in these broad measures employed by means of cognitive abilities, research shows that correlational analyses, and to investigate sex females and males also differ in specific differences in abilities by comparing the abilities, including some that have been performance of males and females on implicated in performance on mechanical mechanical aptitude and on a nonverbal test tasks, such as spatial abilities and perceptual of general intelligence. An exploration of speed (Hedges and Nowell, 1995; Hegarty the interrelationships among some of these and Sims, 1994). abilities by means of the newly revised Beta Studies of sex differences in mechanical III (Kellogg & Morton, 1999) should clarify aptitude have consistently favored males. the basis for the sex differences in For example, research comparing females mechanical ability. The following and males on the Bennett Mechanical hypotheses are postulated: Comprehension Test (BMCT), one of the oldest and most widely used instruments in (1) Men will perform significantly its field, typically indicates that males better than women on the BMCT and outperform females by a wide and WTMA. significant margin (e.g., Fortson, 1991). The

54 Osprey Journal of Ideas and Inquiry (2) Differences between the mean Procedure scores of females and males will be greater The subjects were tested in groups of on the BMCT than on the WTMA. no fewer than two and no more than eight people. Testing sessions lasted for (3) The correlation between the BMCT approximately two hours, with variations in and the WTMA will be statistically length depending on the number of subjects. significant for females and males as well as The three tests were administered to each for the total group. group of subjects in one session, with a 15- minute break in the middle. In the first half (4) The correlation between the Beta III of the session, subjects were asked to read and the two mechanical aptitude tests will be and sign the informed consent and had an smaller than the correlation between the opportunity to ask questions. After this WTMA and the BMCT. informal question and answer session, subjects were asked to fill out the demographic information sheet. Finally, the Methods Beta III was administered. The two mechanical aptitude tests were administered Participants in counterbalanced order, during the second The sample was composed primarily of half of the testing session, after a IS-minute white college students enrolled in various break. Eighteen females and seven males undergraduate psychology courses. The 57 took the BMCT first followed by the subjects tested (36 females and 21 males) WTMA and 15 females and 13 males took were recruited from the UNF Psychology the WTMA first followed by the BMCT. Department Participant Pool. Participation was strictly voluntary. The mean age for females was 24.69, ranging from 18 to 52. Results The mean age for males was 23.48, ranging from 17 to 50. One-way ANOVAs of total scores on the BMCT, WTMA, and Beta III were Materials and Apparatus conducted. The results, seen in Table 1, The study involved the administration show (1) a significant sex difference of three paper-and-pencil tests and a (p<.001) favoring males, on the BMCT; (2) demographic information sheet designed to a significant sex difference (p=.039) gather basic background data. The BMCT favoring males, on the WTMA; and (3) a and the WTMA are group instruments negligible sex difference (p=.605), favoring comprised, respectively, of 68 and 60 items females, on the Beta III. measuring mechanical aptitude. The BMCT The correlation between the BMCT and uses pictures to present mechanical the WTMA for the total group was problems and asks the examinee to choose significant (p<.01), as were the separate from three possible options. The WTMA is correlations between these measures for in the Same format as the BMCT; the both sex groups. These correlations are examinee is asked a question and given shown both jointly and separately in Figures three options from which to choose the 1,2, and 3. correct answer. The intention of the WTMA The correlation between the BMCT and is to minimize sex differences by using test the Beta III for the total group was .462, content that is based on common everyday which is significant at the .01 level. The objects and events. The Beta III is also a corresponding correlation between the group test and it measures nonverbal WTMA and the Beta III was lower (.363) intellectual abilities by means of five distinct but still significant at the .01 level. subtests.

Osprey Journal of Ideas and Inquiry 55 Table 1. Sex Differences on Mechanical Aptitude Measures and Beta III Female Male One-Wa Anova Test Mean SD Mean SD F p BMCT 40.22 7.05 49.05 7.28 20.298 <.001 WTMA 44.14 4.31 46.86 5.26 4.481 .039 Discussion Beta III 104.61 9.61 103.24 9.63 .270 .605 The small sample size is obviously a serious limitation of the study. Another

Figure 1. Corr.lallon between the BMCT and limitation related to the small sample size 6o~ ____~th=e~W~T=M~A~~~r~~~e_~_b_I~~_O~UP~-----, was the small number of males compared to the number of females, which likewise W . influenced the results. Also, though T 50 . M statistical' analysis does not show any order A a •• a'a .0 aa effects, the number of females and males in a the two order conditions (BMCT first vs.

JU~ ______~ ____~ ____~ ____~ WTMA first) is uneven. W • " m _ ro So far, all the major hypotheses have BMCT been confirmed. In addition, there are some interesting trends that will be explored in • 'silnifietat .t the .01 level (two-tailed) future research, as the sample size increases and analyses concerning differences between the sexes are expanded into subtest

Figure 2: Correlation hetween the BMCT and and item data. The results confirmed the the WTMA for Male. expected relationships among the constructs 80 assessed by the BMCT, the WTMA, and the Beta III. As expected, the results show 50 BcoBa a .. D greater convergence between the BMCT and DD D tJ " the WTMA than between the Beta III and of 40 the two mechanical aptitude tests.

30 Also as expected, males scored higher 20 30 40 50 80 70 than females on both of the mechanical

BMO aptitude tests. Males scored significantly higher than females on the BMCT. Even

•• siplificlIlI at the .01 level (two-tail«i) though males outperformed females on the WTMA, the sex difference was smaller, a finding that suggests the WTMA may indeed be a fairer measure of mechanical aptitude for women, provided that the Figure 3: Correlation between t .... BMCT and the WTMA for female. predictive power of both measures is found 8Or------, to be comparable. The fact that the gender W gap was smaller on the WTMA compared to T 50 M the BMCT does support the claim that the A ... T 40 D" WTMA reduces sex differences and is less D biased towards women. No significant sex difference was expected on the Beta III and none was obtained, although females did BMCT score higher than males on that test. The finding that women performed ··significlIlt at the .01 level (two..taiied) better (compared to men) on the WTMA than on the BMCT is promising. However, future research on the predictive validity of the WTMA is needed.

56 Osprey Journal of Ideas and Inquiry Lessons Learned Halpern, D. F. Sex Differences in Cognitive Abilities. (Mahwah, NJ: Lawrence I have become very familiar with the Erlbaum Associates, 1992). library as a result of countless of hours spent researching my subject. I have learned the Hedges, L.V., and Nowell, A "Sex different stages of the research process from Differences in Mental Test Scores, research design to administering group tests Variability, and Numbers of High Scoring to statistical analysis. I have learned the Individuals." Science 269 (1995): 41-45. formalities, procedures, and practicalities that are involved in the research process, Hegarty, M., and Sims, V. K. including ordering materials and writing to "Individual Differences in Mental the IRB just to name a few. However, my Animation During Mechanical Reasoning." knowledge expands beyond the time Memory and Cognition 22 (1994): 411-430. consuming processes of recruiting subjects, collecting data, entering data, and writing Jensen, A R. The G Factor: The informed consent forms. I have gained an Science ofMental Ability. (Westport, CT: insight into how much time, effort, Prager, 1998). dedication, and patience is needed in my future profession. I believe that this Kellogg, C.E., & Morton, N.W. Beta III experience will benefit me tremendously for Manual. (San Antonio, TX: Psychological continuing onto graduate school and Corporation, 1999). eventually becoming an 110 . Tippins, N. T. "Review of the Bennett Mechanical Comprehension Test in D.J. References Keyser and R.C. Sweetland eds., Test Critiques VIII (Austin, TX: Pro-Ed, 1991). Anastasi, A, and Urbina, S. , 7th ed. (Upper Saddle Vernon, P. The Structure ofHuman River, NJ: Prentice Hall, 1997). Abilities. (New York: Wiley, 1950).

Bennett, G. K. Bennett Mechanical Wiesen, 1. P. Wiesen Test of Comprehension Test: Manual. (San Mechanical Aptitude, PAR Edition: Antonio, TX: Psychological Corporation, Professional Manual. (Odessa, FL: 1994). Psychological Assessment Resources, 1999). Chapman, L. K. "An Examination of Mean Score Gender Differences and Adverse Impact for Two Tests of Mechanical Aptitude." (Unpublished Masters Thesis, California State University, Long Beach, 1998).

Fortson, H. D. "An Investigation of Gender Differences in Mechanical Aptitude." (Unpublished Doctoral Dissertation, California School of Professional Psychology, Los Angeles, 1991).

Osprey Journal of Ideas and Inquiry 57