﻿ TPC Journal-Vol 11-Issue-3 - FULL ISSUE – Page 10

# TPC Journal-Vol 11-Issue-3 - FULL ISSUE

270 The Professional Counselor | Volume 11, Issue 3 distribution (Field, 2018); however, the Shapiro-Wilk and Kolmogorov-Smirnov tests are sensitive to large sample sizes and should be interpreted cautiously. In addition, the data should be tested for linearity (Mvududu & Sink, 2013). Furthermore, extreme univariate and multivariate outliers must be identified and dealt with (i.e., removed, transformed, or winsorized; see Field, 2018) before a researcher can proceed with factor analysis. Univariate outliers can be identified via z-scores (> 3.29), box plots, or scatter plots, and multivariate outliers can be discovered by computing Mahalanobis distance (see Field, 2018). Figure 2 Flow Chart for Reviewing Exploratory Factor Analysis Sample Size Assumption Checking Factor Extraction Factor Retention Factor Rotation Naming the Rotated Factors A subjects-to-variables ratio (STV) of 10:1 (at least 10 participants for each test item), or 200 participants, whichever produces a larger sample. 1. Normality: Skewness < ±2 & kurtosis < ±7 2. Inter-Item Correlation Matrix: Every item correlates between r = .20 and r = .80 or .85 with at least 3 other items. 3. Bartlett’s Test of Sphericity: p < .05 4. KMO Test for Sampling Adequacy: ≥ .70 Maximum Likelihood: If the data are largely consistent with a normal distribution (skewness & kurtosis ≤ ±1). Principal Axis Factoring: Moderate violations of normality (skewness 1.1 to 2 & kurtosis 1.1 to 7). Principal Component Analysis: A method of item reduction; not a viable factor extraction method. 1. Kaiser Criterion: Tends to overestimate the number of factors; however, it can be used to extract the initial factor solution. 2. Percentage of Variance Explained by a Factor: ≥ 5% 3. Scree Plot: Graphical representation of factors and corresponding Eigenvalues with a clear bend in the line graph depicting the number of factors to extract. 4. Parallel Analysis: See description in the “Factor Retention” section. Oblique Rotation (e.g., direct oblimin): Use when factors inter-correlate. Orthogonal Rotation (e.g., varimax): Use when factors are uncorrelated. * h2 values .30 to .99, factor loadings ≥ .40, cross-loading ≥ .30 Factor names should be brief (approximately one to four words) and capture the theoretical meaning of the group of items that comprise the factor. Using a research team can enhance the rigor of the factor names.

RkJQdWJsaXNoZXIy NDU5MTM1