| Chapter 2: | Literature Review |
|
|
To test their hypothesis (value congruence increases trust), a general linear model (GLM) was specified as follows:
where Ti is the trust measure for the ith individual, VSi is a binary variable with 1 indicating the ith individual received scenario 1, and 0 indicating the ith individual received scenario 2 or 3, VOi is a binary variable with 1 indicating the ith individual received scenario 3, and 0 indicating the ith individual received scenario 1 or 2, and εi is the error term for the ith individual.
The GLM procedure was used rather than an analysis of variance because each treatment group did not have exactly the same number of subjects. Scenario 2, value neutral organization, was chosen as the baseline to facilitate interpretation of the coefficients. If α1 is positive, then value congruence leads to increased trust. Similarly, if α2 is negative, then value conflict leads to decreased trust. Thus a positive value for α1 and a negative value for α2 would support the prediction of value congruence and conflict having an effect on trust. A comparison of the magnitudes of α1 and α2 will indicate whether the reward for congruence is greater or less than the penalty for value conflict. From the literature, they were not able to specify a hypothesis about the relative magnitudes of α1 and α2. Equation (1) was estimated for each of the four trust questions, as well as the average for questions one through three.
In running their model, they found that value congruence had a positive effect on trust for the matching values group (and a negative impact on trust for the opposing values group) across each dimension of trust and for the average and overall trust levels, indicating that value congruence has both a strong and positive effect for value-positive organizations and a negative effect for value-negative organizations when compared to the value neutral case.