Reshaping cross-section into longitudinal data

[Chiara Burastero]

Good morning,


For my master thesis I have to analyze the impact of a treatment X at time t on an outcome Y at t+1.
Currently, my dataset is cross-sectional having a column identifying only once the individual and all the other variables, collected both at time t and time t+1, not being distinguished by the time element.

ID	Xt	Xt+1
1	a	b
2	c	d
3	e	f

Therefore, I was wondering whether I have to reshape my data in order to have two rows for each individual in the dataset to obtain something like this:

ID	Xt	Xt+1
1	a	.
1	.	b
2	c	.
2	.	d

However, it is not clear to me which commands I should use on STATA to obtain the above results.


Thank you

[Andrew Musau]

"Xt+1" cannot be a Stata variable name. You should start by naming [Xt Xt+1 Xt+2 ...] as [X0 X1 X2 ...]. Then just have an X variable and a time variable.

* Example generated by -dataex-. For more info, type help dataex
clear
input byte ID str1(X0 X1)
1 "a" "b"
2 "c" "d"
3 "e" "f"
end

reshape long X, i(ID) j(time)

Res.:

. l, sepby(ID)

     +---------------+
     | ID   time   X |
     |---------------|
  1. |  1      0   a |
  2. |  1      1   b |
     |---------------|
  3. |  2      0   c |
  4. |  2      1   d |
     |---------------|
  5. |  3      0   e |
  6. |  3      1   f |
     +---------------+

[George Ford]

reshape to long, the Y variable should be a single variable, and identify t+1 as post and mark the treated.

as Andrew demonstrates.