For a research project (RCT) I am working on, I'm trying to identifying how different demographic groups differed in their performance on certain indices over time (like finance, health etc - measured via monthly surveys) and how that may have interacted with treatment groups these participants (randomized into 6 different treatment groups). Therefore, the main DV are four indices and the main IV are certain demographic characteristics from our participant pool.
I had no issues calculating the income effects for all the variables EXCEPT household income (HHInc2019) - which is the total self-reported household income of the participants in the year 2019.
HHINC | Freq. Percent Cum.
--------------------+-----------------------------------
Less than 10,000$ | 3,826 27.26 27.26
10,000$-20,000$ | 3,251 23.16 50.42
20,000$ - 30,000$ | 2,774 19.76 70.18
30,000 - 40,000$ | 1,685 12.00 82.19
40,000$ - 50,000$ | 941 6.70 88.89
50,000$ - 60,000$ | 527 3.75 92.65
60,000$ - 70,000$ | 325 2.32 94.96
70,000$ - 80,000$ | 212 1.51 96.47
80,000$ - 90,000$ | 151 1.08 97.55
90,000$ - 100,000$ | 101 0.72 98.27
100,000$ - 110,000$ | 88 0.63 98.90
110,000$ - 120,000$ | 35 0.25 99.15
120,000$ or above | 120 0.85 100.00
--------------------+-----------------------------------
Total | 14,036 100.00
What I'm stuck on: I want to adjust this variable based on the zip code of the participants but I'm not sure how to do this! I merged my dataset with the ACS survey with median household income by zip code, but how can I now adjust this categorical variable based on what I have?
Alternatively - would it work if I just use the ACS income variable as a control in my interaction effects model?
