By following Dr. Woodridge's advice on Statalist, I applied the ordered probit before I ran the two-stage least square. Here is the link where I got this advice:
https://www.statalist.org/forums/for...=1617356656297
I need more opinions or advice on doing this and how to interpret the coefficient.
z is my continuous instrument variable. Here is my code.
Internet UseFre5 | Freq. Percent Cum.
-------------------------------------+-----------------------------------
0. Never | 25,326 41.04 41.04
1. Occasionally | 3,754 6.08 47.12
2. Several times a week | 5,315 8.61 55.73
3. Once a day | 7,327 11.87 67.60
4. Several times a day\ All the time | 19,994 32.40 100.00
-------------------------------------+-----------------------------------
Total | 61,716 100.00
cheated | Freq. Percent Cum.
---------------+-----------------------------------
-1. Don't Know | 600 0.74 0.74
0. No | 77,288 95.06 95.79
1. Yes | 3,420 4.21 100.00
---------------+-----------------------------------
Total | 81,308 100.00
tabulate Internet_UseFre5, gen(d)
xi: oprobit Internet_UseFre5 z i.year i.country
predict d0h d1h d2h d3h d4h
xi:ivregress 2sls cheated (d2 d3 d4 d5 = d1h d2h d3h d4h) i.year i.country if cheated>-1, robust
d1= Never (ref)
d2= Occasionally
d3= Several times a week
d4= Once a day
d5= Several times a day\ All the time
d0h=Never (ref)
d1h=Occasionally
d2h=Several times a week
d3h=Once a day
d4h=Several times a day\ All the time
Instrumental variables 2SLS regression Number of obs = 61,345
Wald chi2(33) = 484.26
Prob > chi2 = 0.0000
R-squared = .
Root MSE = .43568
------------------------------------------------------------------------------
| Robust
cheated | Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
d2 | 1.088271 1.186184 0.92 0.359 -1.236606 3.413148
d3 | -.9737936 1.018278 -0.96 0.339 -2.969581 1.021994
d4 | .1615734 .2719028 0.59 0.552 -.3713463 .694493
d5 | .1429395 .0656786 2.18 0.030 .0142118 .2716673
_Iyear_2013 | -.0154354 .0048815 -3.16 0.002 -.025003 -.0058678
_Iyear_2014 | -.0005229 .0478025 -0.01 0.991 -.0942141 .0931683
_Icountry_2 | -.0699497 .0767293 -0.91 0.362 -.2203363 .0804369
_Icountry_3 | -.0547166 .045909 -1.19 0.233 -.1446965 .0352634
_Icountry_4 | -.1057375 .0901781 -1.17 0.241 -.2824834 .0710083
_Icountry_5 | -.0951339 .0794164 -1.20 0.231 -.2507872 .0605195
_Icountry_6 | -.0460229 .0340111 -1.35 0.176 -.1126835 .0206377
_Icountry_7 | -.1486605 .0912381 -1.63 0.103 -.3274839 .0301629
_Icountry_8 | -.1216828 .0963118 -1.26 0.206 -.3104505 .0670848
_Icountry_9 | -.1182297 .0943019 -1.25 0.210 -.3030581 .0665986
_Icountry_10 | -.1185502 .091776 -1.29 0.196 -.2984277 .0613274
_Icountry_11 | -.0619843 .0483694 -1.28 0.200 -.1567865 .032818
_Icountry_12 | -.1002474 .0707969 -1.42 0.157 -.2390069 .038512
_Icountry_13 | -.0345833 .0445779 -0.78 0.438 -.1219543 .0527877
_Icountry_14 | -.0467488 .0625823 -0.75 0.455 -.1694078 .0759102
_Icountry_15 | -.0412707 .048214 -0.86 0.392 -.1357683 .0532269
_Icountry_16 | -.1368185 .105492 -1.30 0.195 -.3435791 .0699421
_Icountry_17 | -.1212968 .0871777 -1.39 0.164 -.292162 .0495684
_Icountry_18 | -.0880058 .0668085 -1.32 0.188 -.2189481 .0429364
_Icountry_19 | -.1029886 .1010635 -1.02 0.308 -.3010694 .0950923
_Icountry_20 | -.1101948 .0796603 -1.38 0.167 -.2663262 .0459365
_Icountry_21 | -.0736924 .0706536 -1.04 0.297 -.212171 .0647862
_Icountry_22 | -.0899739 .0817671 -1.10 0.271 -.2502345 .0702866
_Icountry_23 | -.0795511 .0879443 -0.90 0.366 -.2519188 .0928166
_Icountry_24 | -.0439061 .0310109 -1.42 0.157 -.1046864 .0168743
_Icountry_25 | -.1058146 .0697757 -1.52 0.129 -.2425725 .0309433
_Icountry_26 | -.0985241 .0907529 -1.09 0.278 -.2763965 .0793482
_Icountry_27 | -.1403589 .0991263 -1.42 0.157 -.3346429 .053925
_Icountry_28 | -.0896062 .1030279 -0.87 0.384 -.2915372 .1123248
_cons | .0808889 .0816798 0.99 0.322 -.0792007 .2409785
------------------------------------------------------------------------------
Instrumented: d2 d3 d4 d5
Instruments: _Iyear_2013 _Iyear_2014 _Icountry_2 _Icountry_3 _Icountry_4
_Icountry_5 _Icountry_6 _Icountry_7 _Icountry_8 _Icountry_9
_Icountry_10 _Icountry_11 _Icountry_12 _Icountry_13
_Icountry_14 _Icountry_15 _Icountry_16 _Icountry_17
_Icountry_18 _Icountry_19 _Icountry_20 _Icountry_21
_Icountry_22 _Icountry_23 _Icountry_24 _Icountry_25
_Icountry_26 _Icountry_27 _Icountry_28 d1h d2h d3h d4h
If I only do the 2sls, my result is below:
xi:ivregress 2sls cheated (Internet_UseFre5 =z) i.year i.country if cheated>-1, first robust
Instrumental variables 2SLS regression Number of obs = 61,345
Wald chi2(30) = 2085.56
Prob > chi2 = 0.0000
R-squared = .
Root MSE = .19229
----------------------------------------------------------------------------------
| Robust
cheated | Coefficient std. err. z P>|z| [95% conf. interval]
-----------------+----------------------------------------------------------------
Internet_UseFre5 | .0535563 .0079977 6.70 0.000 .0378811 .0692315
_Iyear_2013 | -.014059 .0018248 -7.70 0.000 -.0176356 -.0104824
_Iyear_2014 | .0593533 .0158059 3.76 0.000 .0283744 .0903322
_Icountry_2 | -.0020516 .0071126 -0.29 0.773 -.015992 .0118887
_Icountry_3 | .0013182 .0066464 0.20 0.843 -.0117086 .014345
_Icountry_4 | -.0071633 .0063434 -1.13 0.259 -.0195961 .0052695
_Icountry_5 | -.012271 .0067308 -1.82 0.068 -.025463 .0009211
_Icountry_6 | -.0189606 .0060433 -3.14 0.002 -.0308053 -.0071159
_Icountry_7 | -.0853614 .0098064 -8.70 0.000 -.1045817 -.0661411
_Icountry_8 | -.0291369 .0059988 -4.86 0.000 -.0408943 -.0173794
_Icountry_9 | -.0300534 .0060916 -4.93 0.000 -.0419928 -.018114
_Icountry_10 | -.0390998 .0071333 -5.48 0.000 -.0530808 -.0251188
_Icountry_11 | -.0172367 .0056415 -3.06 0.002 -.0282938 -.0061796
_Icountry_12 | -.0174856 .0059346 -2.95 0.003 -.0291171 -.005854
_Icountry_13 | .0173656 .0067088 2.59 0.010 .0042165 .0305146
_Icountry_14 | .0059199 .0074152 0.80 0.425 -.0086137 .0204534
_Icountry_15 | .0048612 .0065805 0.74 0.460 -.0080364 .0177588
_Icountry_16 | -.0389928 .0063976 -6.09 0.000 -.0515318 -.0264538
_Icountry_17 | -.033172 .0052384 -6.33 0.000 -.0434392 -.0229049
_Icountry_18 | -.0369552 .0091222 -4.05 0.000 -.0548344 -.019076
_Icountry_19 | .0077851 .0075843 1.03 0.305 -.0070798 .02265
_Icountry_20 | -.0619523 .0109204 -5.67 0.000 -.0833559 -.0405488
_Icountry_21 | .009305 .0068316 1.36 0.173 -.0040847 .0226946
_Icountry_22 | .0124935 .007683 1.63 0.104 -.002565 .0275519
_Icountry_23 | .0277552 .0076436 3.63 0.000 .012774 .0427365
_Icountry_24 | -.0110729 .005812 -1.91 0.057 -.0224643 .0003185
_Icountry_25 | -.0369066 .0052607 -7.02 0.000 -.0472174 -.0265958
_Icountry_26 | -.0038788 .0060677 -0.64 0.523 -.0157713 .0080137
_Icountry_27 | -.0684286 .0103549 -6.61 0.000 -.0887238 -.0481334
_Icountry_28 | .0090188 .0066887 1.35 0.178 -.0040909 .0221284
_cons | -.050694 .0170752 -2.97 0.003 -.0841608 -.0172273
----------------------------------------------------------------------------------
Instrumented: Internet_UseFre5
Instruments: _Iyear_2013 _Iyear_2014 _Icountry_2 _Icountry_3 _Icountry_4
_Icountry_5 _Icountry_6 _Icountry_7 _Icountry_8 _Icountry_9
_Icountry_10 _Icountry_11 _Icountry_12 _Icountry_13 _Icountry_14
_Icountry_15 _Icountry_16 _Icountry_17 _Icountry_18 _Icountry_19
_Icountry_20 _Icountry_21 _Icountry_22 _Icountry_23 _Icountry_24
_Icountry_25 _Icountry_26 _Icountry_27 _Icountry_28 z
My question is should I really do the ordered probit before 2sls? Any advice would be wonderful.
https://www.statalist.org/forums/for...=1617356656297
I need more opinions or advice on doing this and how to interpret the coefficient.
z is my continuous instrument variable. Here is my code.
Internet UseFre5 | Freq. Percent Cum.
-------------------------------------+-----------------------------------
0. Never | 25,326 41.04 41.04
1. Occasionally | 3,754 6.08 47.12
2. Several times a week | 5,315 8.61 55.73
3. Once a day | 7,327 11.87 67.60
4. Several times a day\ All the time | 19,994 32.40 100.00
-------------------------------------+-----------------------------------
Total | 61,716 100.00
cheated | Freq. Percent Cum.
---------------+-----------------------------------
-1. Don't Know | 600 0.74 0.74
0. No | 77,288 95.06 95.79
1. Yes | 3,420 4.21 100.00
---------------+-----------------------------------
Total | 81,308 100.00
tabulate Internet_UseFre5, gen(d)
xi: oprobit Internet_UseFre5 z i.year i.country
predict d0h d1h d2h d3h d4h
xi:ivregress 2sls cheated (d2 d3 d4 d5 = d1h d2h d3h d4h) i.year i.country if cheated>-1, robust
d1= Never (ref)
d2= Occasionally
d3= Several times a week
d4= Once a day
d5= Several times a day\ All the time
d0h=Never (ref)
d1h=Occasionally
d2h=Several times a week
d3h=Once a day
d4h=Several times a day\ All the time
Instrumental variables 2SLS regression Number of obs = 61,345
Wald chi2(33) = 484.26
Prob > chi2 = 0.0000
R-squared = .
Root MSE = .43568
------------------------------------------------------------------------------
| Robust
cheated | Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
d2 | 1.088271 1.186184 0.92 0.359 -1.236606 3.413148
d3 | -.9737936 1.018278 -0.96 0.339 -2.969581 1.021994
d4 | .1615734 .2719028 0.59 0.552 -.3713463 .694493
d5 | .1429395 .0656786 2.18 0.030 .0142118 .2716673
_Iyear_2013 | -.0154354 .0048815 -3.16 0.002 -.025003 -.0058678
_Iyear_2014 | -.0005229 .0478025 -0.01 0.991 -.0942141 .0931683
_Icountry_2 | -.0699497 .0767293 -0.91 0.362 -.2203363 .0804369
_Icountry_3 | -.0547166 .045909 -1.19 0.233 -.1446965 .0352634
_Icountry_4 | -.1057375 .0901781 -1.17 0.241 -.2824834 .0710083
_Icountry_5 | -.0951339 .0794164 -1.20 0.231 -.2507872 .0605195
_Icountry_6 | -.0460229 .0340111 -1.35 0.176 -.1126835 .0206377
_Icountry_7 | -.1486605 .0912381 -1.63 0.103 -.3274839 .0301629
_Icountry_8 | -.1216828 .0963118 -1.26 0.206 -.3104505 .0670848
_Icountry_9 | -.1182297 .0943019 -1.25 0.210 -.3030581 .0665986
_Icountry_10 | -.1185502 .091776 -1.29 0.196 -.2984277 .0613274
_Icountry_11 | -.0619843 .0483694 -1.28 0.200 -.1567865 .032818
_Icountry_12 | -.1002474 .0707969 -1.42 0.157 -.2390069 .038512
_Icountry_13 | -.0345833 .0445779 -0.78 0.438 -.1219543 .0527877
_Icountry_14 | -.0467488 .0625823 -0.75 0.455 -.1694078 .0759102
_Icountry_15 | -.0412707 .048214 -0.86 0.392 -.1357683 .0532269
_Icountry_16 | -.1368185 .105492 -1.30 0.195 -.3435791 .0699421
_Icountry_17 | -.1212968 .0871777 -1.39 0.164 -.292162 .0495684
_Icountry_18 | -.0880058 .0668085 -1.32 0.188 -.2189481 .0429364
_Icountry_19 | -.1029886 .1010635 -1.02 0.308 -.3010694 .0950923
_Icountry_20 | -.1101948 .0796603 -1.38 0.167 -.2663262 .0459365
_Icountry_21 | -.0736924 .0706536 -1.04 0.297 -.212171 .0647862
_Icountry_22 | -.0899739 .0817671 -1.10 0.271 -.2502345 .0702866
_Icountry_23 | -.0795511 .0879443 -0.90 0.366 -.2519188 .0928166
_Icountry_24 | -.0439061 .0310109 -1.42 0.157 -.1046864 .0168743
_Icountry_25 | -.1058146 .0697757 -1.52 0.129 -.2425725 .0309433
_Icountry_26 | -.0985241 .0907529 -1.09 0.278 -.2763965 .0793482
_Icountry_27 | -.1403589 .0991263 -1.42 0.157 -.3346429 .053925
_Icountry_28 | -.0896062 .1030279 -0.87 0.384 -.2915372 .1123248
_cons | .0808889 .0816798 0.99 0.322 -.0792007 .2409785
------------------------------------------------------------------------------
Instrumented: d2 d3 d4 d5
Instruments: _Iyear_2013 _Iyear_2014 _Icountry_2 _Icountry_3 _Icountry_4
_Icountry_5 _Icountry_6 _Icountry_7 _Icountry_8 _Icountry_9
_Icountry_10 _Icountry_11 _Icountry_12 _Icountry_13
_Icountry_14 _Icountry_15 _Icountry_16 _Icountry_17
_Icountry_18 _Icountry_19 _Icountry_20 _Icountry_21
_Icountry_22 _Icountry_23 _Icountry_24 _Icountry_25
_Icountry_26 _Icountry_27 _Icountry_28 d1h d2h d3h d4h
If I only do the 2sls, my result is below:
xi:ivregress 2sls cheated (Internet_UseFre5 =z) i.year i.country if cheated>-1, first robust
Instrumental variables 2SLS regression Number of obs = 61,345
Wald chi2(30) = 2085.56
Prob > chi2 = 0.0000
R-squared = .
Root MSE = .19229
----------------------------------------------------------------------------------
| Robust
cheated | Coefficient std. err. z P>|z| [95% conf. interval]
-----------------+----------------------------------------------------------------
Internet_UseFre5 | .0535563 .0079977 6.70 0.000 .0378811 .0692315
_Iyear_2013 | -.014059 .0018248 -7.70 0.000 -.0176356 -.0104824
_Iyear_2014 | .0593533 .0158059 3.76 0.000 .0283744 .0903322
_Icountry_2 | -.0020516 .0071126 -0.29 0.773 -.015992 .0118887
_Icountry_3 | .0013182 .0066464 0.20 0.843 -.0117086 .014345
_Icountry_4 | -.0071633 .0063434 -1.13 0.259 -.0195961 .0052695
_Icountry_5 | -.012271 .0067308 -1.82 0.068 -.025463 .0009211
_Icountry_6 | -.0189606 .0060433 -3.14 0.002 -.0308053 -.0071159
_Icountry_7 | -.0853614 .0098064 -8.70 0.000 -.1045817 -.0661411
_Icountry_8 | -.0291369 .0059988 -4.86 0.000 -.0408943 -.0173794
_Icountry_9 | -.0300534 .0060916 -4.93 0.000 -.0419928 -.018114
_Icountry_10 | -.0390998 .0071333 -5.48 0.000 -.0530808 -.0251188
_Icountry_11 | -.0172367 .0056415 -3.06 0.002 -.0282938 -.0061796
_Icountry_12 | -.0174856 .0059346 -2.95 0.003 -.0291171 -.005854
_Icountry_13 | .0173656 .0067088 2.59 0.010 .0042165 .0305146
_Icountry_14 | .0059199 .0074152 0.80 0.425 -.0086137 .0204534
_Icountry_15 | .0048612 .0065805 0.74 0.460 -.0080364 .0177588
_Icountry_16 | -.0389928 .0063976 -6.09 0.000 -.0515318 -.0264538
_Icountry_17 | -.033172 .0052384 -6.33 0.000 -.0434392 -.0229049
_Icountry_18 | -.0369552 .0091222 -4.05 0.000 -.0548344 -.019076
_Icountry_19 | .0077851 .0075843 1.03 0.305 -.0070798 .02265
_Icountry_20 | -.0619523 .0109204 -5.67 0.000 -.0833559 -.0405488
_Icountry_21 | .009305 .0068316 1.36 0.173 -.0040847 .0226946
_Icountry_22 | .0124935 .007683 1.63 0.104 -.002565 .0275519
_Icountry_23 | .0277552 .0076436 3.63 0.000 .012774 .0427365
_Icountry_24 | -.0110729 .005812 -1.91 0.057 -.0224643 .0003185
_Icountry_25 | -.0369066 .0052607 -7.02 0.000 -.0472174 -.0265958
_Icountry_26 | -.0038788 .0060677 -0.64 0.523 -.0157713 .0080137
_Icountry_27 | -.0684286 .0103549 -6.61 0.000 -.0887238 -.0481334
_Icountry_28 | .0090188 .0066887 1.35 0.178 -.0040909 .0221284
_cons | -.050694 .0170752 -2.97 0.003 -.0841608 -.0172273
----------------------------------------------------------------------------------
Instrumented: Internet_UseFre5
Instruments: _Iyear_2013 _Iyear_2014 _Icountry_2 _Icountry_3 _Icountry_4
_Icountry_5 _Icountry_6 _Icountry_7 _Icountry_8 _Icountry_9
_Icountry_10 _Icountry_11 _Icountry_12 _Icountry_13 _Icountry_14
_Icountry_15 _Icountry_16 _Icountry_17 _Icountry_18 _Icountry_19
_Icountry_20 _Icountry_21 _Icountry_22 _Icountry_23 _Icountry_24
_Icountry_25 _Icountry_26 _Icountry_27 _Icountry_28 z
My question is should I really do the ordered probit before 2sls? Any advice would be wonderful.
