On p. 757, the authors write, \"Internal consistency, assessed using Cronbach\'s

Question

On p. 757, the authors write, "Internal consistency, assessed using Cronbach's , was .76 for the Netherlands, .78 for Sweden, and .79 for Spain." What are they telling you here? Be specific. o 72% ooo T-Mobile Wi-Fi 1:03 PM content.ebscohost.com 1,0800 and were from the and .79 for Spain. We performed a confir- educational sector (45.5 percent) for the most matory factor analysis ECFA) on the perfor- Swedish organizations had, on average, mance items. The unconstrained model of a part 180 employees (SD 179, and were mai nly one factor solution provided reasonable fit to from the retail sector (l0.7 percent). Detailed the data (x 384.11, di 27 description of the samples is provided in the index IGFI 94, comparative fit index ICFI Results section as well as in Table lll. 90, root mean square error of RMSEAJ .08). Constraining the factor load. Measures ings to be invariant between countries, still results in a good fit (x 476.37, df 37, GFI nizational Variables (Level 20 .93, CFI .87, RMSEA .07. Following the recommendations by Cheung and Rensvold The extent o raining as an antecedent vari- 12002) Values between and -0.02 able was identified by HR managers, who were ked to rate separately the percentage (in should be viewed with increasing suspicion, while values great er than -0.02 indicate decimal points of temporary and permanent likely difference between the constrained and ployees who receive training, This resulted unconstrained models" (p. 254- e used n two independent predictors, with each ACFI in the corn parison of the unconstrained scale ranging from 0 percent to 100 percent. model with a model constraining factor Countries were included as dummy vari- loadings to be equivalent between coun- ables for Spain and the Netherlands (coded with a value of 10, with Sweden used as refer- tries. Comparing the Spanish with the Dutch sample, ACFI was .0l for the comparisons ence. The value of 0 represented the resp between the Spanish and Swedish, as well as tive other two countries. between the Swedish and Dutch sam Control Variables .02. Thus, we can conclude that measure- ment invariance between country samples organizational size, measured by the number not perfect but may still be acceptable. of employees, was included as a continuous Control vuriahles at the eml variable. Sectors were included as dummy were tenure (years in the organization, age variables for the retail and food manufactur- continuously), gender (dichoto- (measured ing sectors, with the education sector used as mous variable with 0 for females and 1 for reference, males) and educational level measured with a Likert-scale format using the six-level inter- Employee Variables (Level 1) national educational classification scheme, Work contract type, as a moderator variable, ISCED, OECD, 1999). was dummy-coded, with a value of 0 repre- senting employees with temporary contracts Statistical Analysis and 1 for those with permanent contracts. Multilevel analysis is appropriate if employee Performance the dependent variable data (level 1) is nested within organizations was assessed by employees rating six items Abramis, 1994) related to their performance 0evel 2), as it allows examination of within- subjects" and "between-subjects" variances during their last working week on the follow- separately. We used hierarchical linear model- ing tasks: making decisions, performing with- ing (HIM) by Bryk and Raudenbush (992) out mistakes, devoting themselves to work, and conducted models separately for each achieving their objectives, taking initiative, and taking responsibility. Answers were pro- country with the correspondi ng HLM6R vided using a five-point Likert scale from 1 software (Raudenbush, Bryk, & o 5 (very good) and a self-reported 2005. very bad) performance mean was computed. Internal Analyses for variance ANOVAs) for consistency, assessed using Cronbach's a, selfreport performance Table ll) showed was .76 for the Netherlands, .78 for Sweden, significant

Explanation / Answer

Internal consistency is typically a measure based on the correlations between different items on the same test. Internal consistency is usually measured with Cronbach's alpha, a statistic calculated from the pairwise correlations between items. Coefficient alpha will be negative whenever there is greater within-subject variability than between-subject variability. In this article, Coefficient alpha value for all three variables in between 0.7 to 0.8. Therefore we say that there is a acceptable correlations between variables for all countries.

Related Questions

Navigate

• Browse (All)
• Browse O
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.