I am working on a study about hotels, which analyzes the possible influence of job sastifacción (MSQ, 20 questions of scale 1 to 5) on customer satisfaction (scale 1 to 5, questions that evaluate different departments of the hotels). I have data from 12 hotels and a sample of 65 workers (supervisors from 5 departments) and approximately 14,000 clients, who indicated their satisfaction with each department (5 evaluations) of these hotels. I want to estimate a model of structural equations, but I do not know how to match these data, because the subjects answered different questions (tests).
I have thought of starting from the customer data base and then matching the evaluation of each client with the satisfaction of the evaluated worker.
The problem is that the data is repeated a lot. For example, for hotel 1 there are 1956 responses from customers, but there are only 5 responses from workers. I can match the evaluation of department 1 with the satisfaction of worker 1, but that answer will be repeated 1956 times,. I think this will affect the correlation and, therefore, the model of structural equations.
Is there any way to solve this problem? I have tried to find articles with similar studies, but they do not explain in detail how they made the different samples match. Thanks for your attention. I hope you can help me.
As the sample size is robust two-way ANOVA can be used. But for large data it is important to look at histograms to understand the summary measures. The outliers also need to be addressed. Post this data editing regression will deliver results if the dependent variable is on metric scale.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts