I have a data set as shown in the image attached. the data set is about peak expiratory flow rate (PEFR) of five different locations (Johnsganj, ashoka, katra, alopi & rambagh)in the city with varying level of pollution. These locations are set as dummy variable. PEFR at each locations are surveyed by using peak flow meter. Individuals related health parameters like BP, height, weight etc are recorded.
My questions are:
1. Is this the right arrangement of data for cross - sectional data analysis?
2. if not, what should be the arrangement of data?
3. what are the preliminary steps needed to make the data ready for cross sectional data analysis?
4. How to fit the most optimal regression model for this data set?
Hello Sugandh,
From the data slice shown, I'm presuming that the Johnsganj site is the baseline (so cases would have the pattern, 0,0,0,0 across the four dummy variates).
1. Yes, if your chief aim is to determine whether site helps explain PEFR differences after accounting for the other variables.
2. OK as is, unless I've misunderstood your query.
3. None, really. After fitting the regression model, it's a good idea to evaluate the usual assumptions (linearity of relationships, normality and homoscedasticity of residuals). As well, the data collection must have been undertaken in such a way so as not to cause a lack of independence among observations.
4. "Most optimal," according to what criterion? Explained variance? Parsimony? Information? Something else? I'd start with a model that includes all proposed explanatory variables and work from there. Without knowing your specific research question(s), it's hard to be more specific.
Good luck with your work?
Dear Sugandh Kumar Choudhary
You need to see the data management requirement as per your software.
In stata, the data management for cross sectional analysis simply comprises of the variable arrangement as per the model sequency from Dependent to independent variables
- Badges
- Science topic