Research title
The Effect of Managerial Type of Control on Employee's Affective Commitment in Foodservice Small Business
Research objectives:
Specific research objectives is to examine the impact of owner-manager type of control on the employees’ affective commitment.
Research question:
How particular type of managerial control influence employee's level of affective commitment?
Managerial Type of Controls (Independent Variables) are:
1. Activity Control
2. Outcome control
3. Capability Control
4. Combination Control
The Dependent variable are:
1. Employee's level of affective commitment"
2. Relationship quality between manager and employees
Previously developed scales were used to measure each of the variables of interest.
Activity control, capability control and outcome control were measured by modified scale based on Miao et al. (2007).
Survey Participants indicated their agreement with each item by using a 7-point Likert scale ranging from “strongly disagree” to “strongly agree.”
This also apply to the instrument used to measure the 'Affective commitment".
I believe that a Likert scale that contains seven values - strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree - is ordinal.
Dear Dendy Rosamn
You can use SEM no issue at all. Moreover you can use Regression Analysis with Annova to check the difference between parameters.
Hi Dendy,
as suggested, you can use SEM or multiple regression analysis.
I suggest performing firstly a confirmatory factor analysis in order to test the measurement model.
If the measurement model fits the data, you can specify structural relationships between the latent construct in order to test the effect of the independent variables on the dependent one.
You are right; the 7-point Likert scale is an ordinal variable.
However, as long as the distribution approaches the normality, you can treat
a seven-point scale as a continuous variable. This should not affect the parameters estimation.
Therefore, you can use a maximum likelihood estimation. If you have other dichotomous or categorical variables I suggest weighed least squares.
Best,
Gianmaria
Elaborating on Gianmaria's point: If the distribution on your 7-point scales is not roughly symmetric, I would not use straight-up ML, but either an ordinal-data method such as WLS or an estimator with some robustness to non-normality (e.g., SEs from bootstrap or a sandwich estimator).
- Badges
- Science topic