How to assess a variable plays a role in a model as fixed or random effect?
How can you teach that the same variable plays a role as fixed and random effect in different models with respect to different perspectives? Gender, age, or stg else? Would please give some various examples?
My last research that studying about the effect of teaching method to change the motivativation of student learning show the act of teacher to force that the good result must doing all the exam at home, it will give them high score to reach the opportunity to go abroad and make their parent happiness. That research show that the implication of psychologie forcement have good impact to change the learning method of student behaviof
I would teach this way: A parameter can be a kept constant by the researcher, to study the effect of other variables. For example, if you have a database of people's responses, you can filter by field sex=male to find out what just the males think (as a function of age). Sex is now a constant.This is a better approach than only populating the database with responses from males, and having to try to repeat the survey with females, as time has moved on, and this may change things.
If now you find that there is no significant difference between sexes, you can turn the filter off to pool the results for both sexes, and find out how age affects responses with a bigger sample (male + female).
In the first case, the field sex is a constant, in the second it is a (categorical) variable.
Similarly with age. If you filter your database with age = 13 to 19, then the parameter or field called age is now held constant (age="teenager"). You can then see how just the teenagers respond. If you turn this filter off, age becomes a variable (I would not call it a random variable) and you can report on how many people of any age responded. (The measured heights of the males constitutes an example of a random variable -- that is, it approximates a normal distribution.)
You do not determine the effect of a variable in a model. You have to first assume the probable effect. Therefore, its important to freeze your research design based on what you want to measure . Collect your data on the variables and test your assumptions using the appropriate statistics. It will be difficult to explain any general condition. If you have a specific situation then it will be easier to discuss the methods to measure the effect
- Badges
- Science method
- Similar topics
- Teaching