PRAXIS 3

Praxis 3: Triangulate Your Findings

Praxis 3: Triangulate Your Findings

The measures for organizational culture from an individual firm perspective were adapted. The rationale for using their scales is due to their strong theoretical foundation. Each organizational culture construct (i.e. clan? ad hocracy? hierarchy? and market) consisted of four items and each item represented 4 different organizational dimensions; kind of organization? leadership practices? cohesiveness? and the focal point of the organization's existence. These dimensions have found empirical grounds in the literature. The respondents were asked to express their agreement or disagreement with the items with regard to the similarity of the description to their business. Each organizational culture construct was measured with four items on a seven-point scale (ranging from 1-strongly disagree to 7-strongly agree).

A factor analysis was employed to assess the construct validity of multi-item scales. Since? organizational culture does not form a single empirical dimension? four different factor analyses for the four organizational cultures were used. Separate factor analyses for clan? ad hocracy? and hierarchy cultures identified one factor having an eigenvalue greater than two. All four items in each scale significantly loaded on their representative a priori factor. Factor analysis for market culture? on the other hand? identified two factors. Since? one of the items loaded significantly on the second factor? it was omitted from the data analysis. Internal reliability of each organizational culture and control variables was assessed using the Cronbach's a. All measures met the minimum criterion of 0.60. The composite index for organizational culture was determined by summing the items in the scale that loaded on their latent factor. Table I presents the results of reliability and validity assessments.

The analysis

A binary logit model was used to test the proposed relationships between organizational culture and export activity. The binary logit model was estimated using the maximum-likelihood nonlinear estimation routine. The model tested in this study can be defined in the form of: Equation 1 where the ß n 's represent the coefficients of the four organizational cultures and control variables (X n 's) and P represents the dichotomous export activity (1 - exporter and 0 - non-exporter).

Table II exhibits the results of the logistic regression parameters of the model. The overall fit of the model was significant at ? 2=23.48 and p=0.000274. Overall? the full model classified 81.20 percent of the cases correctly. Also? the model predicted 74.32 percent of non-exporting firms and 59.02 percent of exporting firms correctly.

The proposed associations between the independent and dependent variables were assessed with the sign of the coefficients. The significance of relationships was assessed with the univariate Wald test statistics. The critical value of 2 was used to indicate any significant relationship between the predictor and criterion variables.

The second stage of the analysis included a univariate t-test to determine mean differences between exporting and non-exporting firms. The results indicated some major differences in organizational cultures of two types of firms as presented in Table III.

Results

The findings of this exploratory research suggest four major ...