if we are using a sample where the majority is male e.g., in some professions like engineering, which are male dominant, and mainly young as the whole population of engineers falls b/w 25-45 years, do we need to control age or gender. Moreover, the study design is stratified random sampling, and our analytical technique is SEM, please guide
It always depends on the kind of study you are trying to do and importance of a particular variable in your study. For example:
1. If you are trying to study the behavior of Engineers in petroleum schematics, comparing it with engineers in structural layout planning; Neither age nor gender is relevant. There activities, there stress level, their performance and pre- post- work habits may be few of the important parameters.
2. If you are trying to study the pace of performance of engineers in one particular industry, then age can be one of the variables you may use for understanding the change in performance linked with change in age.
3. If you are trying to understand the empirical responsiveness of engineers based on gender, then you can certain use gender as a variable.
In essence, (in my opinion) gender dominance is not a factor. Percentile is the factor to draft a conclusion.
Some demographic characteristics, such as gender and age, are generally controlled for in almost every statistical model. However, and as Rafeh said, this is normally just a default adjusting and it should depend on what you are trying to answer. If, as you said, age and gender are very homogeneous and you don't expect to see any differences between different ages and genders, you have two options: See if adjusting for it makes any difference (preferred) or bypassing it (also ok). I hope it helps
Thanks Sir Rafeh Jamal, and Sir Nuno Mendonca, for your insightful guidance, Sir Rafeh, I'm studying the impact of interpersonal conflict on job performance and withdrawal behavior, in that case gender, or age are relevant or not
In many social studies gender tends to be irrelevant (public opinion, use of media or information technololgies, etc.), but in my opinion, in this case that you mention it should be a variable to observe. Family conciliation could affect more working women, and also gender inequalities are quite clear in professional status and incomes in many jobs...
Respecting age, allways is a variable to control, you have to analize carefully in order to make the age groups, because there might be important differences to take into account.
Thanks Sir Jose for your valuable guidance, it is quite helpful, thanks again
If you have already measured the variables in question, then testing for their effects is better than just making the assumption that something does not matter.
The male and age are just your assumption. you can use demographic as your variable to see their effect on what you are testing.
- Badges
- Science topic