Page 1 of 1

Time based data

Posted: May 21st, 2019, 12:57 pm
by arkagwa
Dear Josh et al

Is it possible to design a CSPro dictionary to capture data say for a period of 1 year in 3 cohorts (each cohort after 4 months, i.e Jan , May and September). I don't want data to overwrite data, i need to keep data for each cohort intact when i collect again data for the same variables. When i export data should remain as shown in the in the picture below. The aim is to measure the impact of my interventions. How can we achieve this design in CSPro?
20190521_195003.jpg
20190521_195003.jpg (55.6 KiB) Viewed 3093 times

Re: Time based data

Posted: May 23rd, 2019, 4:54 pm
by aaronw
I would be hesitant to send enumerators into the field and have them updating the same cases. Also, I think the application would quickly become overly complicated. I would design my questionnaire, so each case is specific to a single cohort. During analysis you could then link the cohorts together on their shared cohort id + wave id.

Re: Time based data

Posted: May 23rd, 2019, 7:37 pm
by josh
I agree with Aaron that you don't want to use the same data file and risk overwriting the data from the previous round. You could, however, use the data from the previous round as a read-only lookup file to display the data from the earlier rounds in protected fields while entering data from the current round. This could make it easier to link them up in the end.