Summary of Content, Concurrent and Construct Validity of the Leadership Skills Inventory

Leadership Skills Inventory (LSI) Validity

Introduction

The Leadership Skills Inventory (LSI) is designed to help students from upper-elementary to post-secondary levels assess their leadership skills. This study examines the content, concurrent, and construct validity of the LSI, building upon research by Edmunds & Yewchuk (1996).

Background and Need

Importance of Leadership in Gifted Education

Marland's (1972) definition of leadership as a component of giftedness has driven the need for tools to identify potential leaders, evaluate leadership training, and understand leadership as a construct (Feldhusen, 1994; Karnes & Meriweather-Bean, 1991).

Existing Limitations

Despite efforts (Cattell & Cattell, 1975; Renzulli, Hartman & Callahan, 1971), a comprehensive leadership indicator is lacking (Hollingworth, 1939). Existing indicators vary in research basis, with unexplored inter-relationships (Friedman, Friedman & Van Dyke, 1984; Sisk, 1985).

Practical and Theoretical Gaps

A reliable leadership measure is needed in both business (House, 1971; McGregor, 1960) and education (Chemers & Rice, 1973; Hollingworth, 1939). Empirical research on leadership indicators since 1980 is limited.

Prevalent Leadership Indicators

The four main indicators of leadership:

1. Pencil and paper tests (Karnes & Chauvin, 1986).

2. Elections, nominations, or rankings (Kitano & Kirby, 1986).

3. Observed leadership behavior (Friedman et al., 1984).

4. Past leadership behavior (Bass, 1981; Stogdill, 1974).

The Leadership Skills Inventory (LSI)

Description

The LSI (Karnes & Chauvin, 1985) is a 125-item Likert-type measure assessing strengths in nine categories:

1. Fundamentals of Leadership.

2. Written Communication Skills.

3. Speech Communication Skills.

4. Values Clarification.

5. Decision Making Skills.

6. Group Dynamic Skills.

7. Problem Solving Skills.

8. Personal Development Skills.

9. Planning Skills.

Concerns and Prior Research

While Feldhusen (1994) sees merit in the LSI, concerns exist regarding its content, concurrent, and construct validity (Eckart, 1988; Kerr, 1988; Lee, 1988).

Content Validity

Karnes & Chauvin (1985) reported content validity was established through expert and youth reviews.

Concurrent Validity

Karnes and D'llio (1988a) compared student self-ratings with teacher ratings, finding no significant differences with experienced teachers but significant differences with inexperienced teachers. The study suggests the LSI is best used by experienced teachers.

Criterion-Related Validity

Karnes and DTlio (1988b) found that students completing a Leadership Studies Program (Karnes, Meriweather & DTlio, 1987) scored similarly to community leaders, suggesting criterion-related validity. However, this is potentially biased due to the students' recent training.

LSI as a Pre/Post Instrument

Karnes, Meriweather and DTlio (1987) used the LSI to measure the impact of leadership training, attributing increased LSI scores to the program. However, the instrument and training program share the same literature base, raising concerns about independent assessment.

Need for Validation

Further validation of the LSI is needed to measure leadership abilities accurately.

Current Study

Objective

This study provides further information regarding the content, concurrent and construct validity of the LSI.

Method

Sample

90 academically gifted Grade 12 students from an urban magnet high school. The sample consisted of 45 girls and 45 boys. The average grade was $81\%$, and average IQ was $126$. None had prior leadership training.

Leadership Indicators

Actual Leadership Behavior

Based on Bass (1981), Friedman et al. (1984), and Stogdill (1974), students generated creative ideas and demonstrated leadership in groups. This task was a modification of Friedman, Friedman, and Van Dyke's (1984) model, aligning with Renzulli's (1979) giftedness conception.

Creative Component

Students developed innovative topics for Expanded Learning Curriculum (ELC) (Edmunds, 1989) within 10 minutes, scored on fluency, flexibility, and originality (Torrance Test of Creative Thinking Norms Technical Manual, 1974a).

• Fluency Score: Number of topics generated minus duplications.

• Flexibility Score: Number of different categories a student's topics fell under.

• Originality Scores:

  • 2 points for topics suggested by only one student.

  • 1 point for topics suggested by only one other student.

  • No points for topics suggested by three or more students.

Fluency, flexibility, and originality were converted to 'z' scores and summed to produce the creativity score.

Leadership Ability Component

Students, in randomly assigned groups of 4-6, ranked ELC topics by consensus and rated each other on:

• How well they helped get the job done.

• How well they helped the group feel good about working together.

Averages were computed and summed to form the leadership ability score.

Task Commitment Score

Derived from willingness to devote (0-10) hours to ELC preparation. Leadership ability and task commitment were converted to 'z' scores and summed with creativity to produce the actual leadership behavior score.

Nomination Indicators

Based on Friedman et al. (1984), Jarecky (1959), Kitano & Kirby (1986), and Renzulli et al. (1971), self, peer, and teacher nomination scores were obtained by ranking the class list for ELC leadership.

Past Leadership Behavior

Based on Bass (1981), Kitano and Kirby (1986), and Stogdill (1974) students provided a list of past leadership positions, evaluated by professionals.

Leadership Skills Inventory (LSI)

Students completed the LSI (Karnes and Chauvin, 1985), scored per manual instructions. Data were examined, and educational leadership experts reviewed all 125 items.

Results

Content Validity

Three experts in educational leadership conducted a semantic comparison of the 125 LSI items to the leadership literature. They found that all test items elicited responses about an individual's leadership abilities, thus, supporting the content validity of the instrument.

Concurrent Validity

Significant positive relationships existed between LSI scores and actual leadership behavior (r = .341, p < .01) and past leadership behavior (r = .389, p < .01), supporting concurrent validity. Significant positive correlations with nomination scores further supported concurrent validity, but this conclusion is tentative due to varying small group sizes.

Inter-correlations Among LSI Categories

All nine LSI categories had significant correlations with each other ($r = .295$ to .761, p < .01).

Factor Analysis

Factor analysis revealed only one factor within the LSI.

This outcome supports the construct validity, indicating that the LSI measures a single leadership construct, not nine distinct categories. Descriptors of the nine categories appear semantic rather than empirically distinct.

Discussion

The study supports the content, concurrent, and construct validity of the LSI. The LSI measures a component of leadership distinct from other indicators, useful as a leadership measure but not for identifying separate leadership categories. It could be used with other measures for comprehensive leadership identification.

Further research is needed to refine LSI items and improve psychometric qualities, potentially through factor analysis with a larger sample. Research into a comprehensive leadership assessment battery is also warranted to consolidate differing research bases and examine inter-relationships among indicators.