Confirmatory Factor Analysis and Internal Consistency of the Student-life Stress Inventory.

The validity and reliability of the Student-life Stress Inventory, SSI, was studied by analyzing the responses made to it by 381 students who were enrolled in classes at a state university. The confirmatory factor analyses and the analysis of variance were used to compute the validity. The internal consistency was used to determine the SSI's reliability. Previously, no factor analysis was computed on the SSI. However, on the other data, the findings concurred with those previously reported, confirming the reliability of the SSI. The confirmatory factor analyses buttressed the notion that the SSI is a valid measurement instrument in determining college students' stressors, reactions to stressors, and their overall stress index.

The Student-Life Stress Inventory, SSI, (Gadzella, 1991) is an instrument designed to study college students' stressors and their reactions to stressors. The inventory is a self-report, paper and pencil questionnaire consisting of 51 items listed under nine categories. It is based on a theoretical model described by Morris (1990). The model focuses on five types (categories) of stressors (Frustrations, Conflicts, Pressures, Changes, and Self-imposed) and four types (sections) of reactions to stressors (Physiological, Emotional, Behavioral, and Cognitive Appraisal)(*){1}.

Numerous studies have been conducted using the SSI. Some of these studies reported the validity and reliability of this inventory. For instance, concurrent validities for the SSI were reported in 1993 (Gadzella & Guthrie) for 87 students' responses and in 1994 (Gadzella) for 290 students' responses. In each study, students' perceptions of their stress levels (mild, moderate, or severe) and their responses to the items in the SSI were analyzed using analyses of variance. Results showed significant differences among the students' stress levels on the nine categories, the two sections (Stressors, and Reactions to Stressors), and the total stress score.

In 1998, Gadzella, Masten, and Stacks reported significant correlations between the SSI scores and the scores in three other instruments: Inventory of Learning Processes (Schmeck, Ribich, & Ramanaiah, 1977); Test Anxiety Inventory (Spielberger, 1980); and Internality, Powerful Others, and Chance Locus of Control (Leverson, 1981). In these studies, Pearson product-moment correlations showed some significant positive and negative correlations between the scores.

Other studies reported the reliability of the SSI. For instance, in 1991, Gadzella, Fullwood, and Ginther computed the internal consistency coefficients for each of the nine categories and the total inventory values of the SSI for 95 students on 3-week test-retest responses. The correlations for the test-retest responses ranged from .57 (Cognitive Appraisal, a reaction to stressors) to .76 (Emotional, a reaction to stressors). In another study (Gadzella & Guthrie, 1993), Pearson product-moment correlations were computed for 87 students on 3-week test-retest responses. The correlations for the whole inventory were .78 for the total group, .92 for the men, and .72 for the women.

Other studies using the SSI (Gadzella, 1994; Gadzella & Fullwood, 1992; Gadzella, Fullwood, & Tomcala, 1992; Gadzella, Ginther, & Fullwood, 1993) reported differences and patterns between groups (e.g., gender, college status, stress levels, and age). Various statistical methods were used in these studies: analysis of variance (ANOVA), t-tests, Pearson product-moment correlations, and the internal consistencies. The use of above-mentioned statistical methods in analyzing research data concurs with the findings reported by Hoyle (1994). He summarized the most frequently used statistical procedures, in a 20-years period, in the Journal of Consulting and Clinical Psychology, JCCP. Hoyle reported that in the years 1972, 1982, and 1992, published research used primarily statistical methods such as: ANOVA, t-tests, and correlations. Not many of the studies used the multiple regression, factor analysis, or structural modeling. The use of the simpler (or traditional) statistical methods in analyzing tests and research data may be due to what average readers understand and/or what graduate students know how to use and explain when reporting their data.

Aiken, West, Sechrest, Reno, Roediger, Scarr, Kazden, and Sherman (1990) reported, from a survey they conducted, that statistical courses in graduate psychology programs lack the training for more complex and specific methods in analyzing measurement instruments. Due to the more complex and specific research questions that are currently raised, Reiss and Stiller (1992) and Hoyle (1994) pointed out that there is a need to adopt the newer, more sophisticated statistical methods (e.g., structural equation modeling).

The purpose of this study was to report on the validity and reliability of the SSI. In addition to computing the concurrent validity, the internal consistency coefficients and confirmatory factor analyses were employed.

According to Bentler (1998)" measurement models used in structural modeling are confirmatory factor analysis models" (p. 183). In confirmatory factor analysis, the researcher knows which variables are good indicators of the factor(s). With reference to the SSI, there are five categories (Frustrations, Conflicts, Pressures, Changes, and Self-imposed), which describe the types of stressors. There are also four categories (Physiological, Emotional, Behavioral, and Cognitive Appraisal), which describe the reactions to stressors.

The procedure usually used in computing the confirmatory factor analysis is to determine the intercorrelations among the variables. Then, a confirmatory factor analysis is computed (based on models set up on data to be analyzed) to determine how much each variable contributes to each factor and the total measurement instrument.

Method

Subjects: There were 381 students, enrolled in psychology classes at a southwestern state university, who volunteered to respond to the SSI. In this group, there were 120 men, 258 women, and three students who did not report their gender; 120 were freshmen, 75 sophomores, 107 juniors, 65 seniors and graduates, and 14 individuals who did not report their college status. Their ages ranged from 17 to 57 years with a mean of 24.47 years and standard deviation of 8.02.

Procedure: In responding to the SSI, subjects first indicated the perceptions of their stress level (mild, moderate, or severe). Then, subjects responded to the inventory by rating each item in the SSI using a 5-point scale in Liken format with 1 = never, 2 = seldom, 3 = occasionally, 4 = often, 5 = most of the time. The values of the first eight categories are summed and recorded. The values of the last category (Cognitive Appraisal) are first revised, then summed and recorded. The total score of the inventory is the summation of the values for the nine categories.

Results

To perform the confirmatory factor analysis, intercorrelations for the nine categories, two sections (Stressors, and Reactions to Stressors), and the Total SSI were computed. These intercorrelations are summarized in Table 1 along with the means and standard deviations for the nine categories.

Table 1

Interrelationships Among the Categories, Sections, and Total SSI and
Means and Standard Deviations for the Nine Categories

                                       Category
Section Category
                            1         2         3         4
Stressors                 .80(*)    .59(*)    .82(*)    .74(*)
  1. Frustrations
  2, Conflicts            .39(*)
  3. Pressure             .55(*)    .41(*)
  4. Changes              .52(*)    .32(*)    .59(*)
  5. Self-imposed         .36(*)    .31(*)    .48(*)    .34(*)

Reactions to
Stressors                 .49(*)    .38(*)    .54(*)    .51(*)
  6. Physiological        .45(*)    .38(*)    .48(*)    .45(*)
  7. Emotional            .48(*)    .35(*)    .56(*)    .50(*)
  8. Behavioral           .38(*)    .29(*)    .40(*)    .44(*)
  9. Cognitive           -.20(*)   -.21(*)   -.16(*)   -.21(*)

Total                     .68(*)    .52(*)    .72(*)    .67(*)

Mean                     17.65      8.25     14.08      8.11
Standard Deviations       4.25      2.00      3.28      2.75

                                           Category
Section Category
                          5         6         7         8        9
Stressors               .71(*)    .59(*)    .65(*)    .51(*)   -.28(*)
  1. Frustrations
  2, Conflicts
  3. Pressure
  4. Changes
  5. Self-imposed

Reactions to
Stressors               .47(*)    .93(*)    .79(*)    .80(*)   -.14(*)
  6. Physiological      .44(*)
  7. Emotional          .49(*)    .64(*)
  8. Behavioral         .36(*)    .58(*)    .57(*)
  9. Cognitive         -.25(*)   -.25(*)   -.28(*)   -.15(*)

Total                   .63(*)    .86(*)    .80(*)    .74(*)   -.22(*)

Mean                    21.82     31.06     12.17     17. 95    6.02
Standard Deviations      3.89      9.83      3.90      5.11     2.12

(*) p < .001

As shown in Table 1, all correlation coefficients among the categories, sections, and the Total SSI were significant (p [is less than] .001). The highest correlation (r = .64) among the categories was between the scores on Emotional and Physiological (reactions to stressors) and the lowest correlation (r = -.15) was between the scores on Cognitive Appraisal and Behavioral (reactions to stressors). For the whole inventory, the highest correlation (r = .86) was between the scores on Physiological category (a reaction to stressors) and the Total SSI and the lowest correlation (r = -.22) was between the scores on Cognitive Appraisal category (a reaction to stressors) and the Total SSI.

Next, the confirmatory factor analysis was computed. When this type of analysis is conducted, models are set up for the data studied and the goodness-of-fit statistics are used to test the hypothesis. The models set up for this study were the Stressor model, Reactions to Stressors model, and the Total SSI model.

In the first model (Stressors), it was hypothesized that the five categories (Frustrations, Conflicts, Pressures, Changes, Self-imposed) would adequately define the Stressors section (a summary of the relationships among all the variables studied on SSI are displayed in Figure 1 and a summary of the goodness-of-fit analyses are presented in Table 2).

[ILLUSTRATION OMITTED]

Table 2

Goodness-of-Fit Summaries for Sections (Stressors and Reactions to
Stressors), and the Models (Stressors, Reactions to Stressors, and
the Total SSI)

                              Goodness of Fit Summary

Section        df        Model   df    [chi      GFI    AGFI   RMSEA
                                      square]

Stressors      10   533.61(*)     5   9.87       .99    .97     .05

Reactions to    6   416.89(*)     2   5.21       .99    .97     .06
 Stressors

Total          36   1172.83(*)   22   53.37(*)   .97    .95     .06

(*) p < .001

GFI = Goodness-of-Fit Index greater than .90 indicate adequate fit
(Bentler & Bonett, 1980)

AGFI = Adjusted Goodness-of-Fit Index greater than .90 indicates
adequate fit (Bentler & Bonett, 1980)

RMSEA = Root Mean Square Approximation less than .10 indicate adequate
fit (Bentler & Bonett, 1980)

The hypothesis for the first model was supported. The results ([X.sup.2] (10) = 533.61, p [is less than] .0001) showed that the five categories in this model were significantly related to the Stressors section (see Figure 1 and Table 2).

Referring to the data from the goodness-of-fit index, the theoretical model (which is the base s for the Stressors section) and the model from the data for the Stressors section did not differ significantly, [X.sup.2] (5) = 9.87, p [is greater than] .05. Stated differently, it means that the two models were similar. According to Bentler and Bonett (1980), when the Goodness-of-Fit and Adjusted Goodness-of-Fit Indexes are greater than .90, the analyses indicate adequate fit of the models. Also, according to Bentler and Bonett (1980), when the Root Mean Square Error of Approximation is less than. 10, the analysis indicates adequate fit of the models. Data in Table 2 showed that the Goodness-of-Fit Index, GFI, was .99, the Adjusted Goodness-of-Fit Index, AGFI, was .99, and the Root Mean Square Error of Approximation, RMSEA, was .05. That is, all the data from the five categories of the Stressors (section) supported the model on stressors (see Figure 1 and Table 2).

In the second model (Reactions to Stressors), it was hypothesized that the four categories (Physiological, Emotional, Behavioral, and Cognitive Appraisal) would adequately define the Reactions to Stressors section (see Figure 1). The hypothesis was supported. The results showed that the four categories in the second model were significantly related to the Reactions to Stressors section, [X.sup.2] (6) = 416.89, p [is less than] .0001(see Figure 1 and Table 2). In terms of the model goodness-of-fit, the theoretical model (which is the basis for the Reactions to Stressors section) and the model obtained from the data for the Reactions to Stressors section did not differ significantly, [X.sup.2] (2) = 5.21, p [is greater than] .05. Stated simply, the goodness-of-fit analysis showed that the two models (theoretical and data obtained models) were similar. Furthermore, the Goodness-of-Fit Index, GFI, was .99, the Adjusted Goodness-of-Fit Index, AGFI, was .97, and the Root Mean Square Error of Approximation, RMSEA, was .06 (see Table 2). All these indexes supported the model on Reactions to Stressors with the four categories (see Figure 1 and Table 2).

Then, the confirmatory factor analysis for the Total SSI model included two latent factors (Stressors, and Reactions to Stressors) that defined the general factor, Student-life Stress Inventory, SSI (see Figure 1). It was hypothesized that the two factors (Stressors, and Reactions to Stressors) would be related to the SSI general factor. The hypothesis was supported. The results showed that the two factors (Stressors, and Reactions to Stressors) were significantly related to the SSI general factor, [X.sup.2] (36) = 1172.83, p [is less than] .0001 (see Figure 1 and Table 2). In addition, the Goodness-of-Fit Index, GFI, was .97, the Adjusted Goodness-of-Fit Index, AGFI, was .95, and the Root Mean Square Error of Approximation, RMSEA, was .06 (see Table 2). All of the indexes supported the construct validity of the general SSI factor, which was composed of the two factors, Stressors and Reactions to Stressors (see Table 2 and Figure 1).

The values for the confirmatory factor analyses of the Stressors, Reactions to Stressors, and the Total SSI are displayed in Figure 1. In confirmatory factor analysis, numerical values between factor(s) and indicator(s) indicate the factor loadings in terms of beta weights (Schumacker & Lomax, 1996). In the analyses of the SSI, all factor loadings for the categories and sections (Stressors, and Reactions to Stressors) were significant (p [is less than] .05). However, the contributions of the categories to the sections (Stressors, and Reactions to Stressors) varied. For instance, the Frustrations category had the highest factor loading (2.95) to the Stressors section; whereas, the Physiological category had the highest factor loading (7.91) to the Reactions to Stressors section. The factor loadings (5.23) for the two sections (Stressors, and Reactions to Stressors) toward the SSI were the same (see Figure 1).

Concurrent validity was also computed for this group of subjects. An analysis of variance was used to determine the differences among the stress level groups (mild, moderate, and severe) and their responses to the items in nine categories of the SSI. The results (see Table 3) showed that there were significant differences (p [is less than] .0001) among the groups in all nine categories, two sections, and the total stress score.

Table 3

Means, Standard Deviations, and F-Ratios for Groups (Mild n = 73,
Moderate n = 247, and Severe n = 61) on Ratings of Their Stressors
and Reactions to Stressors

Section   Category         Group         M          SD      F (2, 378)

I. Stressors (Total)       Mild        61.36      10.45
                           Moderate    70.02      11.53      55.74(*)
                           Severe      80.80      12.05
          Frustrations     Mild        15.26       3.62
                           Moderate    17.56       3.87      34.14(*)
                           Severe      20.87       4.46
          Conflicts        Mild         7.68       1.78
                           Moderate     8.49       1.91      8.71(*)
                           Severe       9.08       2.35
          Pressures        Mild        11.62       3.24
                           Moderate    14.20       2.83      47.54(*)
                           Severe      16.56       2.99
          Changes          Mild         6.44       1.96
                           Moderate     7.98       2.39      51.11(*)
                           Severe      10.11       2.96
          Self-imposed     Mild        20.36       3.44
                           Moderate    21.80       3.78      12.72(*)
                           Severe      23.66       4.10

II. Reactions to           Mild        57.07      12.71
Stressors (Total)          Moderate    66.37      13.32      56.41(*)
                           Severe      82.62      17.80
          Physiological    Mild        24.82       8.00
                           Moderate    30.83       8.32      46.24(*)
                           Severe      39.51      11.46
          Emotional        Mild         9.67       3.16
                           Moderate    12.06       3.51      48.17(*)
                           Severe      15.59       3.74
          Behavioral       Mild        16.25       4.55
                           Moderate    17.31       4.32      36.78(*)
                           Severe      22.57       6.10
          Cognitive App.   Mild         6.34       2.39
                           Moderate     6.18       1.98      9.75(*)
                           Severe       4.95       2.04

III. Total Inventory       Mild        118.42     19.91
                           Moderate    136.40     20.98      71.72(*)
                           Severe      163.43     26.45

(*) p < .0001

Tukey ad hoc tests were computed for all categories, the two sections (Stressors, and Reactions to Stressors), and the Total SSI. The results showed that (a) in all categories (except the Cognitive Appraisal), the two sections, and Total SSI, the "severe" group scored significantly higher (p [is less than] .001) than the "moderate" and "mild" groups, respectively, and (b) in all categories (except the Cognitive Appraisal and Behavioral), the "moderate" group scored significantly higher (p [is less than] .001) than the "mild" group. Higher scores mean that the group(s) had experienced more stressors and reactions to stressors.

The items in the Cognitive Appraisal category dealt with whether the respondents thought and analyzed (a) how stressful their situations were and (b) whether they used the most effective strategies. The research data on Cognitive Appraisal category showed that the "severe" group reported significantly lower (p [is less than] .001) scores than the "moderate" and "mild" groups, respectively. This would mean that the "severe" group experienced less stress in analyzing their stressful situations and the strategies they used than the "moderate" and "mild" groups, respectively.

On Behavioral and Cognitive Appraisal categories, the post-hoc tests showed that the mean scores for the "mild" and "moderate" groups did not differ significantly. That is, although the mean scores were numerically different, they did not differ significantly statistically.

Internal consistencies were computed for the nine categories, the two sections (Stressors, and Reactions to Stressors), and the Total SSI for men, women, and the total group (see Table 4). The lowest internal consistency was .63 for the total group on Self-imposed stressors and the highest internal consistencies were .92 for (a) the men and the total group on the Frustration stressors and (b) the women and the total group on the Total SSI.

Table 4

Internal Consistencies (Alphas) for Student-life Stress Inventory by
Sections and Categories for Gender and Total Groups

                                             Alpha

Section   Category                           Group

                                    Men      Women     Total
                                  (n=120)   (n=258)   (n=381)

Stressors (Total)                   .92       .91       .92
          Frustrations              .74       .69       .70
          Conflicts                 .68       .67       .68
          Pressures                 .81       .79       .80
          Changes                   .86       .87       .86
          Self-imposed              .64       .64       .63

Reactions to Stressors (Total)      .79       .74       .75
          Physiological             .89       .84       .86
          Emotional                 .83       .82       .82
          Behavioral                .78       .69       .71
          Cognitive App.            .89       .78       .82

Total Inventory                     .90       .92       .92

Discussions and Conclusions

In the present study, various analyses were used to obtain information on the validity and reliability of the Student-life Stress Inventory, SSI. Previous studies have reported the concurrent validity (Gadzella & Guthrie, 1993; Gadzella, 1994; Gadzella, Masten, & Stacks, 1998) using the analysis of variance, and the reliability (Gadzella & Guthrie, 1993; Gadzella, Fullwood, & Ginther, 1991) using Pearson product-moment correlations and the internal consistency for the SSI. The findings in the present study concur with those previously reported.

The data on the concurrent validity analyses on the SSI (Table 3) showed that there were significant differences among stress level groups (mild, moderate, and severe) on the stressors and reactions to stressors they experienced. For instance, individuals with "severe" levels of stress experienced a higher intensity of stressors and exhibited greater reactions to stressors than the individuals whose stress levels are considered to be "mild."

This study also provided information on the internal consistency of the SSI. In the present study, the internal consistencies for 381 subjects were .92 for the total test, .90 for men, and .92 for women.

No previous study reported the factor analysis of the Studen-life Stress Inventory, SSI. The confirmatory factor analysis, presented in this study, provides specific information as to how the nine categories (Frustrations, Conflicts, Pressures, Changes, and Self-imposed, Physiological, Emotional, Behavioral, and Cognitive Appraisal), and two sections (Stressors, and Reactions to Stressors) are related to the SSI as a whole. This information buttresses the position that the inventory is a valid research tool. That is, the data collected with the SSI provide valid information on college students' stressors, reactions to stressors, and their total stress index.

To future researchers, who design various measurement instruments, the suggestion is made to compute the data collected on the instruments but to use the more complex and specific methods, like the factor analysis. The factor analysis, confirmatory factor analysis used in the present study, provided a more comprehensive analysis of the validity of the SSI.

Student-life Stress Inventory

(Gadzella, 1991, Copyrighted)

I. STRESSORS

A. As a student (frustrations):

1. I have experienced frustrations due to delays in reaching my goals.

2. I have experienced daily hassles which affected me in reaching my goals.

3. I have experienced lack of sources (money for auto, books, etc.).

4. I have experienced failures in accomplishing the goals that I set.

5. I have not been accepted socially (became a social outcast).

6. I have experienced dating frustrations.

7. I feel I was denied opportunities in spite of my qualifications.

B. I have experienced conflicts which were:

8. Produced by two or more desirable alternatives.

9. Produced by two or more undesirable alternatives.

10. Produced when a goal had both positive and negative alternatives.

C. I experienced pressures:

11. As a result of competition (on grades, work, relationships with spouse and/or friends).

12. Due to deadlines (papers due, payments to be made, etc.).

13. Due to an overload (attempting too many things at one time).

14. Due to interpersonal relationships (family and/or friends, expectations, work responsibilities).

D. I have experienced (changes):

15. Rapid unpleasant changes.

16. Too many changes occurring at the same time.

17. Change which disrupted my life and/or goals.

E. As a person (self-imposed):

18. I like to compete and win.

19. I like to be noticed and be loved by all.

20. I worry a lot about everything and everybody.

21. I have a tendency to procrastinate (put off things that have to be done).

22. I feel I must find a perfect solution to the problems I undertake.

23. I worry and get anxious about taking tests.

II. REACTIONS TO STRESSORS:

F. During stressful situations, I have experienced the following (physiological):

24. Sweating (sweaty palms, etc.).

25. Stuttering (not being able to speak clearly).

26. Trembling (being nervous, biting fingernails, etc.).

27. Rapid movements (moving quickly, from place to place).

28. Exhaustion (worn out, burned out).

29. Irritable bowels, peptic ulcers, etc.

30. Asthma, bronchial spasm, hyperventilation.

31. Backaches, muscle tightness (cramps), teeth-grinding.

32. Hives, skin itching, allergies.

33. Migraine headaches, hypertension, rapid heartbeat.

34. Arthritis, over-all pains.

35. Viruses, cold, flu.

36. Weight loss (can't eat).

37. Weight gain (eat a lot).

G. When under stressful situations, I have experienced (emotional):

38. Fear, anxiety, worry.

39. Anger.

40. Guilt.

41. Grief, depression.

H. When under stressful situations, I have (behavioral):

42. Cried.

43. Abused others (verbally and/or physically).

44. Abused self (used drugs, etc.).

45. Smoked excessively.

46. Was irritable towards others.

47. Attempted suicide.

48. Used defense mechanisms.

49. Separated myself from others.

I. With reference to stressful situations, I have (cognitive appraisal):

50. Thought about and analyzed how stressful the situations were.

51. Thought and analyzed whether the strategies I used were most effective.

References

Aiken, L. S., West, S. G., Sechrest, L., Reno, R. R., Roediger, H. L. III, Scarr, S., Kazdin, A. E., & Sherman, S. J. (1990). Graduate training in statistics, methodology, and measurement in psychology: A survey of PHD programs in North America. American Psychologist, 45, 721-734.

Bentler, P. M. (1998). EQS for Windows (Version 5.7) (Computer software). Los Angles: BMDP Statistical Software, Inc.

Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness-of-fit in the analysis of covariance structures. Psychological Bulletin, 88, 588-606.

Gadzella, B. M. (1991). Student-life Stress Inventory. Copyright, Commerce, Texas, Author.

Gadzella, B. M. (1994). Student-life Stress Inventory: Identification of and reaction to stressors. Psychological Reports, 74, 395-490.

Gadzella, B. M., & Fullwood, H. L. (1992). Differences among university student age groups on their perceptions of stress. Proceedings of the Texas Academy of Science, 95th Annual Meeting, Wichita Falls, TX. Pp. 176-180.

Gadzella, B. M., & Fullwood, H. L., & Ginther, D. W. (1991). Student-life Stress Inventory. Paper presented at the Texas Psychological Convention, San Antonio, TX (ERIC 350 345).

Gadzella, B. M., & Fullwood, H. L., & Tomcala, M. (1992). Students' stressors and reactions to stressors. Paper presented at the Southwestern Psychological Association Convention, Austin, TX.

Gadzella, B. M., Ginther, D. W., & Fullwood, H. L. (1993). Patterns of relationships among types of stressors. Paper presented at the Southwestern Psychological Association Convention, Corpus Christi, TX.

Gadzella, B. M., & Guthrie, D. (1993). Analysis of stress inventory. Proceedings of the Texas Academy of Science, 96th Annual Meeting. University of North Texas, Denton, TX. Pp. 413-431.

Gadzella, B. M., Masten, W. G., Stacks, J. (1998). Students' stress and their learning strategies, test anxiety, and attributions. College Student Journal, 32, 416-422.

Hoyle, R. H. (1994). Introduction to the special section: Structural equation modeling in clinical research. Journal of Consulting and Clinical Psychology, 62, 427-428.

Leverson, H. (1981). Differentiating among internality, powerful others, and change. In Research with locus of control. Vol. 1: Assessment methods. New York: Academic Press. Pp. 15-65.

Morris, C. G. (1990). Contemporary psychology and effective behavior (7th edition). Glenview, IL.: Scott & Foresman.

Reis, H. T., & Stiller, J. (1992). Publication trends in JPSP. A three-decade review. Personality and Social Psychology Bulletin, 18, 46-472.

Schmeck, R. R., Ribich, F. D., & Ramanaiah, N. V. (1977). Development of a self-report inventory for assessing individual differences in learning processes. Applied Psychological Measurement, 1, 413-431.

Schumacker, R. E., & Lomax, R. G. (1996). A beginner's guide to structural equation modeling. New Jersey: Lawrance Erlbaum Associates.

Spielberger, C. D. (1980). Test Anxiety Inventory. CA: Consulting Psychological Press, Inc.

Bernadette M. Gadzella, Ph.D., Professor, Department on Psychology and Special Education. Mustafa Baloglu, Ph.D., Assistant Professor, Department of Counseling, Texas A&M University-Commerce.

Correspondence concerning this article should be addressed to Dr. Bernadette M. Gadzella, Department of Psychology & Special Education, Texas A&M University, Commerce, TX 75429.