Measurement modelIn order to assess the measurement model, exploratory factoranalysis (EFA) was conducted to assess the unidimensionality of itemsand constructs using principal components analysis with a promaxrotation.11 The EFA included all the items with the number of factorsspecified to be the number of latent constructs. The initial factormatrix indicated that all items loaded above 0.60 [17] on the intendedfactor except for one item from strategic view of warehouse. This itemwas dropped and a confirmatory factor analysis (CFA) was conductedto examine the measurement model. The CFA model fit was assessedin terms of four measures identified by authorities: root mean squareerror of approximation (RMSEA), standardized root mean squareresidual (SRMR), Tucker Lewis index (TLI or NNFI), and the adjustedgoodness-of-fit index (AGFI) [30,80]. The recommended target valuesfor each of these fit statistics and the corresponding values for thestudy measurement model are presented in Table 2.In addition to model fit, the reliability, convergent validity, anddiscriminant validity was assessed. As indicated in Table 3, allestimated standard loadings were significant (pb0.05), suggestinggood convergent validity. Also, the average variance extracted (AVE)for each construct exceeded 0.50, demonstrating convergent validity[27]. Finally, the square root of AVE was examined to see if they werelarger than the correlation coefficient to assess discriminant validity[27]. All the correlations exceeded the recommended criterion,indicating discriminant validity.Cronbach alpha values were used to assess the reliability of theitems associated with the latent constructs. According to Gefen et al.[30], Cronbach alphas for constructs in exploratory studies shouldexceed 0.60, and in confirmatory studies they should exceed 0.70.Other authorities suggest that a reliability of at least 0.6 suffices forearly stages of research [64]. The internal consistency of constructs inthis study surpassed the suggested guidelines for exploratory studies.