The Learning Style Inventory

Convergent validity Study in an Applied Career Setting

Patricia A. Cross & Penny Faulkner

Abstract

Kolb's Learning Style Inventory (1978; 1985) has been widely used in educational and occupational settings since the 1970s. Mixed results, evaluating the psychometric properties of the four learning modes and learning types, have been reported. While the instrument is widely used in career development settings, very few studies have examined the relationship between learning styles and on-the-job learning behaviour. The current study evaluated the relationship of the LSI learning types, of 66 employees selected for a federal Public Service executive development program, with supervisors' ratings of observed on-the-job learning behaviours and global evaluations of learning potential. In addition, the relationship of these learning styles to self-reported personality data from the Jackson Personality Inventory, Revised (JPI-R, 1994) was also evaluated. Results indicated that the learning styles are not related to supervisors' ratings of on-the-job learning behaviours, but there is a significant association with supervisors' ratings of overall learning potential(LP) - the Converging and Diverging learners received significantly higher on-the-job global potential ratings than did the Accommodating and Assimilating types. An association between Kolb's learning styles and one JPI construct emerged, with Assimilating types obtaining higher scores for cooperativeness than Converging and Diverging types.

Introduction

Kolb's (1984) experiential learning theory postulates the existence of four learning modes:

• Concrete Experiencing (CE)
• Abstract Conceptualizing (AC)
• Reflective Observation (RO)
• Active Experimentation (AE)

These modes combine to form two learning dimensions: Concrete/Abstract and Active/Reflective. The theory states that while almost every individual utilizes all learning modes to some extent, each person has a preferred learning style, determined by obtaining scores on these two dimensions and mapping them on a grid. The result is four learning styles:

• Divergent (CE/RO)
• Assimilative (AC/RO)
• Convergent (AC/AE)
• Accommodative (AE/CE)

An individual's learning style, Kolb contends, is relatively stable over time. Kolb's Learning Style Inventory (LSI, 1978), revised in 1985, purports to categorize individuals on the basis of their self-reported preferred learning style. Numerous studies have examined the validity and reliability of this instrument, with mixed results (e.g., some more recent publications include Garner, 2000; Lam, 1997; Lam, 1998; Loo, 1999;Yahya, 1998). Very few studies, however, have made use of non-student samples. In fact, we found only three: Lam, 1997; Lam, 1998; and Truluck & Courtenay, 1999.

The present study examined the relationship between learning styles and on-the-job learning behaviours for a group of Canadian Federal Public Service employees screened into a training program for potential executives. The data were derived from instruments that are routinely used in this particular selection-for-training process.

• We were interested in assessing whether their learning styles were related to supervisors' ratings of on-the-job learning behaviours and global evaluations of learning potential.
• We also wanted to investigate whether there was a relationship between learning styles and personality, as measured by the Jackson Personality Inventory, Revised (JPI-R, 1994).
• If Kolb's learning styles reflect stable and real differences in individuals' learning preferences, this ought to be reflected in supervisors' assessments of learning in the work environment. Such a finding would lend support to the validity of Kolb's theory of learning styles and the LSI (1985) itself.
• If specific learning styles are related to different personality characteristics, this would also lend support to the contention that Kolb's learning styles do reflect real and stable individual characteristics.
• If one or more of these styles are positively associated with learning on the job, this information could assist in the preparation of plans in training programs by capitalizing on the most advantageous styles.

Method

A sample of N=66 Canadian Federal Public Service employees was obtained. These individuals had successfully competed for inclusion into an executive development programme. During the competition process, they had voluntarily completed the JPI-R (1994). They were informed at that time that their scores on this inventory would not be used in the selection process. Upon entering the programme, they filled out the LSI (1985) as part of a developmental exercise. As well, each individual provided the names of two recent supervisors, who were contacted by telephone and asked to provide ratings on a series of eight on-the-job learning behaviours and one global evaluation of learning potential (LP) (see Figure 3). Based on their LSI scores, each participant was categorized as either an Assimilator, a Converger, a Diverger, or an Accommodator.

Results

Of the 66 participants, we obtained

• 19 Assimilators
• 21 Convergers
• 14 Divergers
• 12 Accommodators

Rating of Overall Learning Potential:

• one-way between subjects ANOVA on the learning styles and LP scores was significant, F (3, 62) = 4.49, p = .006.
• Convergers (M = 4.80) and Divergers (M = 4.80) had higher ratings than Assimilators (M = 4.50) and Accommodators (M = 4.33)

Ratings of Eight Indicators of On-the-Job Learning:

• one-way between-subjects ANOVAs conducted for each of the eight indicators of on-the-job learning were non-significant

Relationship with JPI-R Scores: The JPI-R is a 300-item True/False personality test.

• oneway ANOVAs indicated that respondents' scores were not related to their learning styles, with the exception of Cooperativeness, F (3, 58) = 3.91, p = .013.
• Assimilators (M = 9.53) scored higher on Cooperativeness than Convergers (M = 5.95) and Divergers (M = 6.67).
• the Cooperativeness scale denotes sensitivity and responsiveness to social pressures and norms; high scorers tend to adopt willingly to the group's views regarding people, places, and events
• hard to interpret these findings because Assimilators are supposed to be less focussed on people and more interested in abstract ideas and concepts; they should be the group least likely to conform to group pressures, given their focus on ideas versus people

Discussion

While LSI scores did not predict scores for on-the-job learning, they did predict participants' overall ratings of learning potential (LP). Specifically, Convergers and Divergers received higher LP ratings than Assimilators and Accommodators. These findings are not inconsistent with Kolb's theory. Convergers are good at finding practical uses for ideas and theories; such a predilection would seem to lend itself readily to the tasks faced by employees aspiring to executive management positions; but, according to Kolb, Divergers prefer to observe rather than take action, which does not seem, on the face of it, to be a particular advantage to employees in this environment. One might speculate that an explanation for the Divergers' high LP ratings may be found in the fact that they tend to prefer to work in groups. This may explain their high LP ratings from supervisors, since teamwork is a highly-valued competency in the Federal Government context.

As for the two groups that received lower ratings of learning potential, Assimilators' lower overall LP ratings might be attributed to the fact that they tend to be more interested in concepts than people, whereas the job required of them likely requires well-developed interpersonal sensitivities and skills. In turn, Accommodators may have received low ratings due to their own preference for "hands-on" work versus the ability to take a more strategic role, a skill that is required in middle-management positions, and that becomes increasingly important as one progresses upward.

The finding that Assimilators obtain higher scores on the JPI-R scale of Cooperativeness does not, on the face of it, appear to support the construct validity of the LSI, since Assimilators ought to be particularly resistant to pressures to conform. Assimilators are supposedly more interested in ideas and concepts than people; thus they ought to be relatively unaffected by a desire to comply with the wishes and opinions of others. The relationship between conformity and this learning style may be more complex than it would appear. It is possible, for example, that since Assimilators are more concerned with ideas than with people, they express agreement with others' views simply because this is the interpersonal path of least resistance, requiring the smallest degree of interaction and discussion.

In sum, we found preliminary, qualified support for the convergent validity of the LSI and its constructs. While the LSI learning styles were associated with supervisors' ratings of learning potential in appropriate ways, the finding that Assimilators self-report as being more conforming than others requires further investigation. Follow-up studies and replication of the present findings are necessary before the utility of the LSI as a predictive tool, and its relationship to other personality constructs, can be determined.