Hi, I'm conducting a longitudinal study to see how certain variables change among participants in four cities in a given period.
However, for some inevitable reason, group data in one city has to be collected ten days later than the other three groups. Since the interval was set as one month, namely 30 days, I wonder whether this will make the data from different cities cannot be merged.
If so, is there anything I can do to address this problem and merge the data?
Thank you in advance!
Depends. Please elaborate on what is your research question, what variables are you monitoring (including contextual data that you control for), and briefly describe your participants. The general rule of thumb is that if any of your independent variables and confounders vary too much, in a non-random manner in those 10 days, it may lead to bias. If, for example, you measure impact of a social media campaign (starting at the same time), 10 days are very meaningful; if you are measuring how often people visit a public space in the follow-up of a redesign, the lag may not matter. In any case, be transparent about it in your paper, and don't freak out if you find out the 10 days matter and you do not have the means to redo the experiment. Just factor in known modifiers to your result (e.g., if weather conditions is a variable, you can obtain data for both periods to develop a factor), make educated guesses to known, but not quantifiable modifiers, and include in your discussion the limitations. Try to have separate conclusions for your more rigorous results and for your wider results.
Thanks for your answer, Viktor!
Yes, I'm researching the effect of a policy on adolescents' mental health. By the time I started to collect the data, the policy had been implemented for a month.
Fortunately, soon after raising this question, we successfully started to collect the last group data. So it's four days later, rather than ten days, than the other group data.
I wonder whether in this situation bias will be produced.
What's more, your advice is really helpful to me, I'll report the time lag.
- Badges
- Science topic
- Similar topics
- Data Mining
- Classification