Is "time analysis" possible here?

Hi! I am a bit at a loss here ... I am conducting a study about the CoronaVirus and how people perceive this. I have 300 participants, and 6 "survey-times", where they have to rate how they feel about their mental health (DV) and how threatened they feel (CV).

Now, one way to do the statistics would be to do an ANOVA with repeated measurements and just compare the averages from time 1 to time 6. But I would prefer to calculate a trend with time analysis. Is this even possible with only 6 times of measurement? And if it is, does anyone have an idea how to do it in R? Or does anyone have an article that explains this, or knows about a similar experiment? I am grateful for every help, I really am lost.

Thank you!


Active Member
Yes, you can study the time effects using a panel-data model. The following framework must work in your case:

DV_ij = α_i + β_1 * Time_j + β_2 * Time_j^2 + γ_1 * X_i1 + ... + γ_p * X_ip + ε,

where DV_ij is the dependent variable for participant i and survey time j. You can consider two specifications: α_i's are fixed effects and α_i's are random effects. You can choose the better specification using Hausman test, AIC and/or BIC... Functions lmer(), plm(), phtest() and glm() in R allow for much of this.
Last edited: