collinearity in DID

[Dion Nuridzin]

I have research data consisting of:

• Treatment: 1 intervention, 0 control
• Time: 0 pre, 1 post1, 2 post2
• Id: Respondent Id, total respondents are 60 with 3 time points resulting in 3*60 IDs.
• Knowledge: knowledge score

The data entries are attached.

When I analyze using DID, I get the following error:

didregress (Pengetahuan_Istri) (Treatment), group(Id) time(Time)
note: Treatment omitted because of collinearity.
model is not identified
The treatment variable Treatment was omitted because of collinearity.
r(198);


I also tried xtset, but the error remains:
xtset Id Time
Panel variable: Id (strongly balanced)
Time variable: Time, 0 to 2
Delta: 1 unit

xtdidregress (Pengetahuan_Istri) (Treatment), group(Id) time(Time)
note: Treatment omitted because of collinearity.
model is not identified
The treatment variable Treatment was omitted because of collinearity.
r(198);

Do you have any suggestions on what I should do?

[George Ford]

Is the treatment the same for all units (as in the data you present)?

If treatment indicates treatment in any period (always 1 or 0, indicating the treament or control groups), you need to create a variable that is 0 in the pre-treatment period and 1 in the treatment period for the treated, and 0 for all controls.

[Dion Nuridzin]

My study design is as follows:

Control Pre-->Post1-->Post2

Intervention Pre-->give intervention -->Post1-->Post2

So, I created the data entries as follows:

• Treatment: 1 for intervention, 0 for control
• Time: there are 3 measurement points (0 pre, 1 post1, 2 post2) (attached).

I still do not understand how to implement your suggestion. Could you please clarify?

Id    Pengetahuan_Istri    Treatment    Time
1    10    Intervensi    Pre
1    13    Intervensi    Pos1
1    14    Intervensi    Post2
2    12    Intervensi    Pre
2    15    Intervensi    Pos1
2    17    Intervensi    Post2
3    11    Intervensi    Pre
3    16    Intervensi    Pos1
3    17    Intervensi    Post2
4    11    Intervensi    Pre
4    14    Intervensi    Pos1
4    16    Intervensi    Post2
5    11    Intervensi    Pre
5    16    Intervensi    Pos1
5    17    Intervensi    Post2
6    11    Intervensi    Pre
6    15    Intervensi    Pos1
6    15    Intervensi    Post2
7    11    Intervensi    Pre
7    14    Intervensi    Pos1
7    17    Intervensi    Post2
8    12    Intervensi    Pre
8    11    Intervensi    Pos1
8    18    Intervensi    Post2
9    9    Intervensi    Pre
9    8    Intervensi    Pos1
9    13    Intervensi    Post2
10    5    Intervensi    Pre
10    9    Intervensi    Pos1
10    13    Intervensi    Post2
11    6    Intervensi    Pre
11    6    Intervensi    Pos1
11    11    Intervensi    Post2
12    6    Intervensi    Pre
12    11    Intervensi    Pos1
12    12    Intervensi    Post2
13    9    Intervensi    Pre
13    14    Intervensi    Pos1
13    15    Intervensi    Post2
14    10    Intervensi    Pre
14    15    Intervensi    Pos1
14    17    Intervensi    Post2
15    10    Intervensi    Pre
15    15    Intervensi    Pos1
15    16    Intervensi    Post2
16    10    Intervensi    Pre
16    16    Intervensi    Pos1
16    16    Intervensi    Post2
17    9    Intervensi    Pre
17    13    Intervensi    Pos1
17    14    Intervensi    Post2
18    7    Intervensi    Pre
18    13    Intervensi    Pos1
18    20    Intervensi    Post2
19    10    Intervensi    Pre
19    6    Intervensi    Pos1
19    14    Intervensi    Post2
20    13    Intervensi    Pre
20    9    Intervensi    Pos1
20    15    Intervensi    Post2
21    10    Intervensi    Pre
21    9    Intervensi    Pos1
21    13    Intervensi    Post2
22    10    Intervensi    Pre
22    7    Intervensi    Pos1
22    10    Intervensi    Post2
23    8    Intervensi    Pre
23    12    Intervensi    Pos1
23    15    Intervensi    Post2
24    8    Intervensi    Pre
24    8    Intervensi    Pos1
24    15    Intervensi    Post2
25    7    Intervensi    Pre
25    5    Intervensi    Pos1
25    13    Intervensi    Post2
26    11    Intervensi    Pre
26    9    Intervensi    Pos1
26    16    Intervensi    Post2
27    11    Intervensi    Pre
27    8    Intervensi    Pos1
27    14    Intervensi    Post2
28    7    Intervensi    Pre
28    7    Intervensi    Pos1
28    16    Intervensi    Post2
29    6    Intervensi    Pre
29    9    Intervensi    Pos1
29    13    Intervensi    Post2
30    10    Intervensi    Pre
30    7    Intervensi    Pos1
30    13    Intervensi    Post2
31    10    Kontrol    Pre
31    10    Kontrol    Pos1
31    10    Kontrol    Post2
32    12    Kontrol    Pre
32    12    Kontrol    Pos1
32    12    Kontrol    Post2
33    11    Kontrol    Pre
33    11    Kontrol    Pos1
33    11    Kontrol    Post2
34    10    Kontrol    Pre
34    11    Kontrol    Pos1
34    11    Kontrol    Post2
35    10    Kontrol    Pre
35    10    Kontrol    Pos1
35    10    Kontrol    Post2
36    8    Kontrol    Pre
36    10    Kontrol    Pos1
36    10    Kontrol    Post2
37    11    Kontrol    Pre
37    10    Kontrol    Pos1
37    10    Kontrol    Post2
38    12    Kontrol    Pre
38    12    Kontrol    Pos1
38    12    Kontrol    Post2
39    19    Kontrol    Pre
39    10    Kontrol    Pos1
39    10    Kontrol    Post2
40    4    Kontrol    Pre
40    5    Kontrol    Pos1
40    5    Kontrol    Post2
41    4    Kontrol    Pre
41    5    Kontrol    Pos1
41    5    Kontrol    Post2
42    5    Kontrol    Pre
42    5    Kontrol    Pos1
42    5    Kontrol    Post2
43    10    Kontrol    Pre
43    10    Kontrol    Pos1
43    10    Kontrol    Post2
44    9    Kontrol    Pre
44    9    Kontrol    Pos1
44    9    Kontrol    Post2
45    10    Kontrol    Pre
45    12    Kontrol    Pos1
45    12    Kontrol    Post2
46    10    Kontrol    Pre
46    10    Kontrol    Pos1
46    10    Kontrol    Post2
47    9    Kontrol    Pre
47    8    Kontrol    Pos1
47    8    Kontrol    Post2
48    6    Kontrol    Pre
48    7    Kontrol    Pos1
48    7    Kontrol    Post2
49    10    Kontrol    Pre
49    10    Kontrol    Pos1
49    10    Kontrol    Post2
50    12    Kontrol    Pre
50    12    Kontrol    Pos1
50    12    Kontrol    Post2
51    8    Kontrol    Pre
51    10    Kontrol    Pos1
51    10    Kontrol    Post2
52    9    Kontrol    Pre
52    8    Kontrol    Pos1
52    8    Kontrol    Post2
53    9    Kontrol    Pre
53    9    Kontrol    Pos1
53    9    Kontrol    Post2
54    10    Kontrol    Pre
54    8    Kontrol    Pos1
54    8    Kontrol    Post2
55    8    Kontrol    Pre
55    8    Kontrol    Pos1
55    8    Kontrol    Post2
56    11    Kontrol    Pre
56    11    Kontrol    Pos1
56    11    Kontrol    Post2
57    11    Kontrol    Pre
57    11    Kontrol    Pos1
57    11    Kontrol    Post2
58    8    Kontrol    Pre
58    10    Kontrol    Pos1
58    10    Kontrol    Post2
59    5    Kontrol    Pre
59    8    Kontrol    Pos1
59    8    Kontrol    Post2
60    9    Kontrol    Pre
60    12    Kontrol    Pos1
60    12    Kontrol    Post2

[George Ford]

Look up dataex and the FAQs here.

You can't get 2 treatment coefficients with xtdidregress, so you'll have to do it with regression if you want both.

egen period = group(time)
replace period = 0 if time == "Pre"
g treat = treatment == "Intervensi"
g did = treat & period!=0

didregress (pengetahuan_istri) (did), group(id) time(period)
reghdfe pengetahuan_istri did , absorb(id period)
reghdfe pengetahuan_istri c.did#i.period , absorb(id period)