Of course, as far as I know, it is to correct the data, then reduce the correlation and then analyze it.
(1) Repeated measurement data is very common in medical and sociological fields, such as observing the curative effect of a group of people before taking medicine, after taking medicine 1 month and 2 months after taking medicine; Another example is that people are divided into different treatment groups, and each group is observed at different time points, and so on.
(2) It is understood here that a group of people have different time points.
(1) Repeated measurement is different from repeated investigation. Repeated measurement is the measurement of "the same population" at different time points.
(2) Repeated surveys are surveys of "different groups" at different time points. For example, the nutritional diet survey is repeated every few years in China, and each survey is not necessarily the same group (of course, there may be repeated people, but generally not too many). This kind of repeated survey data analyzes the changes of some phenomena such as year and generation, which can usually be analyzed by age+period+generation.
? For the same population (not grouped), four time points were measured, and these four time points were used as four independent groups for routine analysis of variance.
? For the same population (not grouped), four time points were measured, and these four time points were taken as random blocks for random block variance analysis.
? Four time points were observed in each group, and t test was conducted at each time point.
(1) The above examples all make a mistake: the non-independence of repeated measurement data is not considered.
(2) The traditional T-test or ANOVA requires the data to meet the precondition of independence. Obviously, repeated measurement data does not meet this prerequisite, and using t-test or variance analysis to deal with repeated measurement data will often increase false positive errors.
There are three main methods: multilevel model, generalized estimation equation and repeated measurement variance analysis.