Description and Development HistoryResearch StudyNEUROPSYCHOLOGICAL ASSESSMENT LABPAUL JONES, ED.D.
UNIVERSITY OF NEVADA, LAS VEGAS
____________________________________________________________________
ADAPTED PERSONALITY SCALE: CogStyle
Assessment Lab Report 2-5
December/1998
This report includes data for one of three instruments designed with specific intent to provide equity for use with both the sighted population and persons with visual disability. CogStyle, has roots in a research project focused on strategies to motivate students with visual disability toward careers in mathematics, science, and technology (Jones, 1996). That project, funded by the Department of Energy, was conducted in 1994-95 After obtaining a
____________________________________________________________________
CogStyle Description and Development
CogStyle is based on a model (Lowen, 1982) originally intended to facilitate computer modeling of personality traits. The four preference factors measured are: attitude (extraversion- introversion); approach (data-ideas); focus (things-people); and process (detailed-contextual).
The four traits measured in CogStyle appear comparable to the dimensions in the Myers-Briggs Type Indicator (Myers & McCaulley, 1985). They are also, with the exception of omission of an anxiety scale, consistent with traits suggested by the so-called "big five" (Costa & McCrae, 1992) model of personality. Prediger, Swaney, and Mau (1993) suggest that the six vocational personality dimensions in Holland's premier model for vocational personality assessment (Holland, Fritzsche, & Powell, 1994) can be interpreted with a two-factor solution comparable to the approach and focus scales noted above.
CogStyle was developed with a goal to create an effective instrument for use by sighted persons and persons with visual disability. Item selection and tryout began with sighted samples with additional field testing then completed using two primary samples of high school students with visual disability and additional data from samples of adults with visual disability.
Development of the 35-item CogStyle started with a 56-item pretest using forced choice comparisons of adjectives, active verbs, occupations, and abilities associated with each of the six Holland dimensions. This pretest and a form of the Myers-Briggs Type Indicator (MBTI) were administered in the Spring of 1995 to 130 sighted subjects drawn from a community college and a university located in the western U.S. With MBTI scores as the primary criteria for item selection, 45 items were chosen for the second stage of field testing.
The new research instrument was then administered to a sample of adults with visual disability and to secondary school subjects drawn from two settings: a cohort of students with visual disability enrolled in a regular comprehensive high school and an additional group of students enrolled in a residential school for students with visual disability. These students completed the research instrument and Holland's Self-Directed Search (Holland, Fritzsche, & Powell, 1994).
The 35 items which comprise CogStyle were selected using data from both sighted subjects and subjects with visual disability. In final form, one adjective and one action verb were selected to represent each of the Holland dimensions. Paired adjectives and then action verbs for each of the six dimensions are presented to the subject with instructions to choose the one which "fits you the best". There are thus 30 paired comparisons. Three additional items to enhance discrimination in the attitude (extraversion-introversion) and two additional items to enhance discrimination in the process (detailed-contextual) factors were selected.
The scoring procedure combines formulae suggested by Prediger, Swaney, and Mau (1993) for scaling of the approach and focus scales, and an application of Bayesian scaling (Jones, 1989) for the attitude and process scales. Scores are on a scale of 1 to 100 and identified as either strong or mixed preference.
In addition to general utility, the CogStyle test is included in this context with growing evidence of a brain-behavior relationship associated with personality traits. Wilson and Languis (1989) found a relationship between introversion-extraversion and brain electrical activity patterns. In a scale comparable to CogStyle focus, Dunn, VanCleave, and Hymes (1984) found evidence suggesting a relationship with closure speed. With a different instrument a trait comparable to CogStyle focus is identified as corteria (Russell & Karol, 1994) and is associated with quick reaction time, high alpha wave interruption in EEG, and other physical signs of high cortical activation level. A prior study in this lab found significant relationships between a focus scale and a variety of cognitive processing tasks.
Method
Participants
Details about the total of 71 participants and general procedure in the series two studies are available in a related report. Six protocols with evidence of random responses on the CogAttention scale were eliminated before proceeding with the study below. CogAttention was the first scale in which participants were asked to respond to auditory stimuli and use a combination of mouse and keyboard responses. The resulting sample of 65 appears comparable to the original group on all other factors (e.g. demographics and performance on other cognitive scales). The eliminated protocols thus appear most likely to be an artifact of the change in response set.
Instrumentation
Descriptions of the instruments used in the series two studies are also available in a related report. The ANAM/TWB cognitive processing scales have a low ceiling on the accuracy score with participants typically obtaining scores of 90% or above. For these scales, particularly with nonclinical samples, the cognitive efficiency score is identified (Levinson & Reeves, 1997) as the more useful score. Cognitive efficiency, originally labeled as "throughput", is a function of both accuracy and speed (correct responses/minute). The efficiency score is the unit of analysis on the cognitive processing scales in the study below.
CogAttention and CogMemory are two instruments developed along with CogStyle for equity in appraisal of persons with and without visual disability. Details about these instruments are available in a related series two report. On these two scales the raw score is the unit for analysis in the study below.
As described above, CogStyle factors are on a scale ranging from 1 to 100. For each of the CogStyle scales, scores of 1-24 and scores of 76-100 are identified as strong preference with scores from 25-75 identified as mixed preference. For example, on the attitude preference (extraversion vs. introversion), a score of 20 would be identified as a strong introversion preference, a score of 78 would be identified as a strong extraversion preference, and a score of 50 would be identified as a mixed preference.
As a part of the demographic information gathering procedure, each participant was presented with a social situation stimulus which is frequently used to identify vocational personality codes. This self-ranking was used to construct a three-letter Holland code for each subject.
Results
Table 1 below displays descriptive statistics including interrelationships among the four CogStyle scales for the participants in this study. The evident overall "people" preference on the focus scale is typical for teacher education students, particularly with a primarily female sample. The extent of the evident overall "context" preference on the process scale, however, is not as expected from previous studies with comparable groups. The especially high correlations of that scale with the other CogStyle preference factors are also not consistent with other data. The direction of the relationship (attention to detail and introversion, attention to detail and preference for working with data, attention to detail and preference for working with things) is not surprising, but the degree of relationship would not have been predicted from other data.
| attitude m=53.7 s.d.=33.73 | approach m=52.0 s.d.=24.39 |
focus m=45.0 s.d.=28.92 | process m=43.7 s.d.=35.32 | ||||
|---|---|---|---|---|---|---|---|
| attitude (extravert-introvert) | 1.00 | -.13 | -.18 | -.30* | |||
| approach (data-ideas) | -.13 | 1.00 | .18 | .70* | |||
| focus (things-people) | -.18 | .18 | 1.00 | .43* | |||
| process (detail-context) | -.30* | .70* | .43* | 1.00 |
Table 2 below examines gender differences in preference reports on the CogStyle scales. Although the difference was not statistically significant, there was an evident trend toward the "people" preference among females on the focus scale. This is consistent with data from comparable instruments. For example, the distribution of male preferences on the MBTI T-F scale, is approximately 70%-30% (Myers & McCaulley, 1985).
| female n=45 | male n=20 | ||||||
|---|---|---|---|---|---|---|---|
| mean (s.d.) | mean (s.d.) | t | p | ||||
| attitude | 53.3 (32.31) | 54.6 (37.61) | .14 | .888 | |||
| approach | 52.2 (23.79) | 51.8 (26.30) | .05 | .957 | |||
| focus | 40.7 (27.89) | 54.9 (29.49) | 1.86 | .068 | |||
| process | 41.7 (33.31) | 46.1 (40.25) | .45 | .649 |
____________________________________________________________________attitude: extravert-introvert
approach: data-ideas
focus: things-people
process: detailed-contextual
Table 3 below provides product moment correlation coefficients among the CogStyle scales and cognitive performance measures used in the series two studies. The attitude (extraversion-introversion) scale was not significantly related to other variables. Two significant relationships were evident for the approach (data-ideas) scale. The "ideas" preference was related to higher efficiency scores on two of the reaction time scales. Two significant relationships were also evident for the focus (things-people) scale. A "people" preference was related to higher performance on one of the reaction time scales. A "things" preference was related to higher performance on the Matching to Sample scale. One significant relationship was found in the process (detail-context) scale with higher performance on one of the reaction time scales associated with the "context" preference.
| 2-choice | react-1 | react-2 | math | spatial | match | cogat | cogmem | |
|---|---|---|---|---|---|---|---|---|
| attitude (extravert-introvert) | -.09 | -.05 | -.08 | -.02 | -.12 | -.10 | -.11 | -.02 |
| approach (data-ideas) | -.30* | -.20 | -.26* | -.02 | -.02 | -.18 | -.17 | .15 |
| focus (things-people) | -.10 | -.26* | .05 | -.04 | .12 | .26* | .16 | .14 |
| process (context-detail) | -.28* | -.08 | -.04 | .08 | .01 | -.02 | -.06 | .11 |
____________________________________________________________________* significant at .05 level
2-choice: Simple Two-Choice Reaction Time Efficiency Score
react-1: Procedural Reaction Time Efficiency Score- Standard Stimuli
react-2: Procedural Reaction Time Efficiency Score- Degraded Stimuli
math: Mathematical Processing Efficiency Score
spatial: Spatial Processing Efficiency Score- Simultaneous
match: Matching To Sample Efficiency Score
cogat: CogAttention Total Raw Score
cogmem: CogMemory Total Raw Score
Tables 4-7 below provides additional data regarding the relationship between CogStyle and the cognitive performance scales used in the series two studies. For this analysis, strong preference groups were identified and contrasted for each of the CogStyle scales. As expected from the correlation coefficients, there were no significant differences in Table 4 describing the strong preference attitude (extravert-introvert) groups. Differences in the strong preference approach (data-ideas) groups described in Table 5 were near the significant level on two reaction time tests and the memory scale, but none reached the .05 level.
In Table 6 the difference in performance of the focus (things-people) groups on the Matching to Sample scale was statistically significant. A possible extraneous influence warranted consideration because of the disproportionate number of males with "things" preference and a significant gender difference (t(63) = 3.81, p < .01) on the Matching to Sample scale. Males scored significantly higher on this scale. The small number of cell entries precluded use of two-way ANOVA, so a t-test comparing the two preference groups was calculated, selecting female students only. With the female sample, the results confirmed the higher level of Matching to Sample performance (t(43) = 2.74, p < .01) by participants with the "things" preference on the CogStyle focus scale.
Table 7 displays differences in the strong preference process (detail-context) groups. A significant difference in favor of the "context" group was evident on the simple 2-choice reaction test. Other differences were not statistically significant.
| strong introvert preference n=23 | strong extravert preference n=23 | ||||||
|---|---|---|---|---|---|---|---|
| mean (s.d.) | mean (s.d.) | t | p | ||||
| 2-choice | 133.4 (17.02) | 128.8 (18.87) | .85 | .398 | |||
| react-1 | 108.0 (8.49) | 106.2 (10.08) | .65 | .522 | |||
| react-2 | 103.2 (9.83) | 98.3 (14.97) | 1.31 | .197 | |||
| math | 24.0 (6.46) | 23.4 (5.24) | .34 | .739 | |||
| spatial | 33.3 (8.45) | 30.7 (5.87) | 1.22 | .230 | |||
| match | 34.7 (10.60) | 33.1 (9.10) | .56 | .579 | |||
| cogat | 15.0 (3.84) | 13.6 (3.70) | 1.29 | .204 | |||
| cogmem | 35.3 (7.97) | 35.4 (6.97) | .03 | .969 |
____________________________________________________________________
| strong ideas preference n=11 | strong data preference n=11 | ||||||
|---|---|---|---|---|---|---|---|
| mean (s.d.) | mean (s.d.) | t | p | ||||
| 2-choice | 137.9 (13.64) | 124.6 (18.24) | 1.93 | .068 | |||
| react-1 | 107.2 (8.08) | 100.9 (11.02) | 1.53 | .141 | |||
| react-2 | 103.3 (10.54) | 94.3 (11.32) | 1.93 | .068 | |||
| math | 24.6 (6.80) | 22.6 (4.93) | .76 | .456 | |||
| spatial | 29.8 (6.62) | 31.2 (6.18) | .52 | .609 | |||
| match | 36.1 (7.69) | 31.2 (6.28) | 1.62 | .122 | |||
| cogat | 15.4 (4.46) | 13.5 (3.56) | 1.06 | .303 | |||
| cogmem | 34.7 (5.27) | 40.6 (8.61) | 1.94 | .066 |
| strong people preference n=17 | strong things preference n=11 | ||||||
|---|---|---|---|---|---|---|---|
| mean (s.d.) | mean (s.d.) | t | p | ||||
| 2-choice | 135.2 (11.27) | 130.2 (14.98) | 1.01 | .323 | |||
| react-1 | 110.1 (9.64) | 107.3 (8.82) | .78 | .443 | |||
| react-2 | 102.2 (13.62) | 103.6 (14.31) | .27 | .789 | |||
| math | 22.6 (5.79) | 22.6 (5.62) | .03 | .979 | |||
| spatial | 28.8 (9.47) | 31.7 (5.51) | .93 | .359 | |||
| match | 29.1 (8.16) | 36.2 (7.66) | 2.29 | .030* | |||
| cogat | 12.9 (3.82) | 13.3 (3.77) | .23 | .823 | |||
| cogmem | 34.8 (6.34) | 36.5 (9.52) | .54 | .590 |
* significant at .05 level
____________________________________________________________________
| strong context preference n=26 | strong detail preference n=18 | ||||||
|---|---|---|---|---|---|---|---|
| mean (s.d.) | mean (s.d.) | t | p | ||||
| 2-choice | 135.3 (13.66) | 123.5 (21.66) | 2.23 | .031* | |||
| react-1 | 105.3 (10.66) | 105.1 (9.18) | .06 | .956 | |||
| react-2 | 100.4 (14.32) | 98.0 (12.67) | .57 | .573 | |||
| math | 23.6 (5.70) | 25.0 (6.26) | .75 | .459 | |||
| spatial | 31.3 (7.78) | 31.2 (6.75) | .01 | .991 | |||
| match | 33.9 (8.65) | 33.8 (8.13) | .02 | .981 | |||
| cogat | 13.0 (3.83) | 12.7 (3.91) | .27 | .791 | |||
| cogmem | 34.1 (6.21) | 37.5 (9.48) | 1.43 | .159 |
* significant at .05 level
____________________________________________________________________
A different view of the CogStyle preference scores was available using the estimated Holland vocational codes obtained through the social simulation exercise gathered as part of the demographic data. Participants were asked to assume that they would have to spend an extended amount of time with one of six groups of people, and rank their first, second, and third choices of groups with whom they would enjoy interacting. Descriptions of the groups were prepared to be consistent with distinct characteristics associated with each of the Holland scales.
Table 8 below displays the mean scores on the CogStyle scales associated with those selections (e.g. 40 participants selected the Realistic group as their first, second, or third choice). There were some anomalies in the selections as compared to actual data on the Holland codes, particularly the large number of participants who chose the Realistic group. Also, the number of participants selecting the Conventional group as first, second, or third was quite small.
The data in Table 8, however, are interesting in the extent to which hypothesized relationships between the CogStyle scales and vocational preferences were in fact evident. The highest mean score on the extraversion-introversion scale was obtained by those who selected Enterprising. The highest mean score on the data-things scale and the detail-context scale was obtained by participants selecting Conventional. A "people" preference was most evident in the Social and Enterprising selections.
| attitude (extravert- introvert) | approach (data- ideas) |
focus (things- people) | process (detail- context) | ||||
|---|---|---|---|---|---|---|---|
| Realistic (n=40) | 52.3 | 46.7 | 49.4 | 39.1 | |||
| Investigative (n=27) | 47.4 | 51.6 | 49.4 | 44.7 | |||
| Artistic (n=34) | 52.1 | 48.3 | 45.3 | 41.5 | |||
| Social (n=55) | 56.6 | 52.3 | 43.2 | 43.0 | |||
| Enterprising (n=31) | 59.4 | 60.0 | 38.4 | 44.1 | |||
| Conventional (n=8) | 46.6 | 63.1 | 46.4 | 60.8 |
Summary and Discussion
In general, the results with the CogStyle instrument are as would be expected with this sample. The attitude (extravert-introvert) and approach (data-ideas) mean scores were near the scale midpoint of 50. On the focus (things-people) scale, the participants were more likely to express a preference toward the "people" pole of this scale, a finding typical for teacher education students predominantly female. Additional study is needed to explore whether the extraordinarily high correlations of the process (detail-context) scale with other CogStyle scales is an artifact of this sample or is instead indicative of need for further refinement of that scale.
As described above, a Bayesian scaling procedure is used with two CogStyle scales, attitude and process. Data in this sample suggest that this scaling appears more likely to produce "strong" preference scores. Forty-six attitude scores and 42 process scores were identified as strong preference. The corresponding numbers for the approach and focus scales were 22 and 28, respectively. Studies with additional sample groups are needed to determine whether this outcome is associated with actual degree of preference or is a function of the scaling system.
A continuing pattern of relationship between personality traits and cognitive performance was evident in this study. The evident difference in the focus (things-people) groups on the Matching to Sample scale was also found in a series one study with that scale using different instrument for personality trait assessment. The series one studies included trait assessment only for the approach and focus factors. An earlier study (Jones, 1994), found some relationship between cognitive performance and a factor comparable to the CogStyle process scale.
The relationship between the focus scale and cognitive performance was evident on several cognitive scales in the series one study, leading to an hypothesis that this personality scale could be tapping a general cortical alertness characteristic. The results of this study provide only modest support for the hypothesis with significant difference evident on only one scale. The difference, though, was in the hypothesized direction associated with cognitive efficiency. Differences between the accuracy scores of the two groups on the Matching to Sample scale were not significant (t(26) = .59, p > .05).
More study is needed to clarify the relationship between personality factors and performance on cognitive tasks. This study, consistent with findings from previous research, supports routinely including some form of trait level personality assessment in applications of clinical neuropsychology.
References
Costa, P.T., Jr., &; McCrae, R.R. (1992). NEO PI-R: Professional manual. Odessa, FL: Psychological Assessment Resources.
Dunn, B.R., VanCleave, R., &; Hymes, D. (1984). The relation of the Myers-Briggs Type Indicator to closure speed. Journal of Psychological Type, 8, 45-47.
Holland, J.L, Fritzsche, B.A., & Powell, A.B. (1994). Technical Manual: Self-Directed Search, Odessa, FL: Psychological Assessment Resources.
Jones, W.P. (1989). A proposal for the use of Bayesian probabilities in neuropsychological assessment. Neuropsychology, 3, 17-22.
Jones, W.P. (1994). Personality correlates of the ANAM battery. Archives of Clinical Neuropsychology, 9, 145-146.
Jones, W.P. (1996). Career assessment for patients with visual disability: Consideration of instrument distortion. Journal of Vision Rehabilitation, 10, 2-7.
Levinson, D.M., & Reeves, D.L. (1997). Monitoring recovery from traumatic brain injury using Automated Neuropsychological Assessment Metrics (ANAM V1.0). Archives of Clinical Neuropsychology, 12, 155-166.
Lowen, W. (1982). Dichotomies of the mind: A systems science model of the mind and personality. New York: John Wiley.
Myers, I.B., & McCaulley, M.H. (1985). Manual: A guide to the development and use of the Myers-Briggs Type Indicator. Palo Alto, CA: Consulting Psychologists Press.
Prediger, D., Swaney, K., & Mau, W. (1993). Extending Holland's hexagon: Procedures, counseling applications, and research. Journal of Counseling & Development, 71, 422-428.
Russell, M.T., &; Karol, D.L. (1994). The 16PF fifth edition Administrator's Manual. Champaign, IL: IPAT
Wilson, M.A., &; Languis, M.L. (1989). Differences in brain electrical activity patterns between introverted and extraverted adults. Journal of Psychological Type, 18, 14-23.
________________________________________________________________________________________________________________________________________