Research Paper Graduate 857 words

Leadership Effectiveness Inventory: Reliability & Regression Analysis

~5 min read

Abstract

This paper presents a quantitative analysis of a Leadership Effectiveness Inventory administered to 472 respondents across four reporting groups: self, direct reports, peers, and supervisors. Using data from 75 questions mapped to nine leadership scales, the study evaluates scale reliability, explores whether items can be dropped without compromising reliability, and conducts regression analysis to identify effective criterion variables. The analysis confirms that most scales meet acceptable reliability thresholds, that reduced-form scales are largely equivalent to originals, and that an effectiveness index constructed from selected questions correlates strongly with key criterion variables — though not uniformly across all items.

📝 How to Write This Type of Paper Writing guide — click to expand

▼

What makes this paper effective

The paper integrates multiple quantitative methods — descriptive statistics, reliability analysis, scale reduction, and regression — into a coherent, sequential analytical narrative.
Data are presented in clearly referenced tables, and the prose explicitly interprets each table's findings rather than leaving numbers to stand alone.
The discussion of reduced-form scales is practically grounded: it links statistical findings to a real-world benefit (gathering equivalent data in one-third less time).

Key academic technique demonstrated

The paper demonstrates item-to-total correlation analysis as a principled method for scale reduction. By dropping the lowest-correlating third of items and re-testing reliability, the author shows how to maintain psychometric validity while improving survey efficiency — a technique central to applied measurement research.

Structure breakdown

The paper opens by describing data collection and sources, then moves through descriptive statistics, scale-level reliability coefficients, a structured item-dropping procedure for reduced scales, and finally a regression analysis using an effectiveness index. Each section builds directly on the prior one, following a conventional quantitative research report format appropriate for a graduate-level methods course or applied organizational research paper.

Introduction and Data Sources

After collecting data from the Nordic questionnaires, a reliability analysis was carried out to ascertain whether the nine scales used could predict overall satisfaction with the manager as a leader. The analysis found the scales reliable, and it was apparent that reducing the number of questions in the questionnaires would not compromise data reliability. In addition, communication, planning/execution, and teamwork were identified as the best predictors of satisfaction.

For this study, data consisted of responses to 75 questions posed to 472 respondents regarding managerial behavior within the organization. Respondents were drawn from four reporting groups: self, direct reports, peers, and supervisors (Table 1).

Descriptive Statistics and Reporting Group Differences

Descriptive statistics were recorded for individual items with descending means and respective standard deviations (Table 2). Examining these results, three variables had means greater than 4.0, indicating high satisfaction: Q69, Q3, and Q5. Conversely, six variables had means below 3.0, indicating dissatisfaction: Q41, Q71, Q31, Q37, Q52, and Q33.

Minor differences in means were observed across the four data sources, but with no systematic pattern among them. The dependent variable — Q75, "Overall, how satisfied are you with this person as a leader?" — recorded an overall mean of 3.92. Table 3 shows the means for this variable broken down by reporting group:

Scale Reliability Analysis

There is a statistically significant difference among these means (F = 2.85, df = 3, 468, p < 0.05, beta-squared = 0.02). Post-hoc testing revealed that the only specific significant difference is supervisors' tendency to rate managers higher than managers rate themselves, by approximately half a point.

Table 4 presents the nine leadership scales derived from 74 individual variables, along with their means, standard deviations, and Cronbach's alpha reliability coefficients:

Two scales — Managing Resources and Communication — have reliabilities slightly below the recommended alpha level of 0.70, though not to the point of being unacceptable. The remaining seven scales are all reliable. Although statistically significant differences between the means of the four reporting groups were found on all scales except Communication, these differences showed no systematic tendencies. Post-hoc tests similarly revealed no consistent pattern of significant differences. When alpha coefficients are calculated separately for each of the four reporting groups, they differ from the full-sample coefficients by no more than 0.02 to 0.05.

3 Locked Sections · 350 words remaining

Reduced-Form Scales · 160 words

"Item-dropping procedure and revised reliability results"

Regression Analysis and Effectiveness Index · 130 words

"Criterion variable correlations and effectiveness index construction"

Conclusions · 60 words

"Summary of key reliability and regression findings"

You’re 40% through this paper. Sign up to read the remaining 3 sections.

130,000+ paper examples AI writing assistant Citation generator Cancel anytime

Key Concepts in This Paper

Scale Reliability Cronbach Alpha Item Reduction Leadership Satisfaction Reporting Groups Effectiveness Index Regression Analysis Managerial Behavior Descriptive Statistics Nordic Questionnaire