Dear Statlist Community,
I am nearing the end of a project and had been hopping of running to ordinal regression model within my study. I am unsure however, what is the recommended best practise for presenting these results. I was hoping to run two different types of ordinal regression, one with a Likert type scale dependent variable and the second with a dichotomous dependent variable. I am using panel data with a fixed effect model on both of these regression
Can I simply utilise the odd's ratios? I had originally hoped to include predicated probabilities but I am (1) struggling to find the commands within stata to generate these and (2) I am worried that with my regression using the Likert scale I will have to generate a lot of information for readers to digest as the main relationship I am examining is age categories. Any help would be greatly appreciated. I provide data and the code used below
1) For my Likert scale dependent variable regression I am utilising the user generate feologit command
2) For my second regression (looks at the relationship between self-perceived employability and actual job gain in time t+1) I use the following syntax:
I am nearing the end of a project and had been hopping of running to ordinal regression model within my study. I am unsure however, what is the recommended best practise for presenting these results. I was hoping to run two different types of ordinal regression, one with a Likert type scale dependent variable and the second with a dichotomous dependent variable. I am using panel data with a fixed effect model on both of these regression
Can I simply utilise the odd's ratios? I had originally hoped to include predicated probabilities but I am (1) struggling to find the commands within stata to generate these and (2) I am worried that with my regression using the Likert scale I will have to generate a lot of information for readers to digest as the main relationship I am examining is age categories. Any help would be greatly appreciated. I provide data and the code used below
1) For my Likert scale dependent variable regression I am utilising the user generate feologit command
Code:
feologit self_percp_nor i.age_cat
Code:
xtlogit jobgain_tp1 i.self_percp_nor, fe or
Code:
* Example generated by -dataex-. For more info, type help dataex clear input double _nomem_encr int _Year float(jobgain_tp1 self_percp_nor age_cat ned_unemp_rate) 800009 2018 0 . 4 3.8 800009 2019 . . 4 3.4 800073 2018 0 . 4 3.8 800100 2017 0 4 1 4.9 800125 2008 1 4 2 3.7 800372 2016 1 4 1 6 800387 2008 1 5 2 3.7 800500 2017 0 . 4 4.9 800540 2010 1 4 2 5 800540 2014 1 2 3 7.4 800671 2017 0 . 4 4.9 800988 2017 1 . 2 4.9 800994 2011 0 2 4 5 800994 2012 0 2 4 5.8 800994 2013 1 2 4 7.3 801303 2008 0 2 2 3.7 801303 2009 0 3 2 4.4 801303 2010 0 7 2 5 801361 2018 1 . 2 3.8 801486 2015 0 4 2 6.9 801486 2016 1 . 2 6 801550 2017 0 4 2 4.9 801599 2008 0 1 3 3.7 801599 2010 0 2 3 5 801635 2013 0 2 2 7.3 801635 2014 0 1 2 7.4 801776 2010 1 6 1 5 801849 2018 0 1 4 3.8 801849 2019 . 1 4 3.4 801858 2016 0 2 3 6 802032 2008 0 2 4 3.7 802032 2009 0 1 4 4.4 802032 2010 0 1 4 5 802032 2011 . . 4 5 802264 2015 1 6 1 6.9 802297 2010 0 6 1 5 802297 2015 0 4 1 6.9 802297 2016 1 6 1 6 802587 2017 0 . 4 4.9 802587 2018 . . 4 3.8 802597 2016 1 6 1 6 802761 2012 0 5 1 5.8 802761 2013 0 5 1 7.3 802761 2014 0 4 1 7.4 802761 2015 0 3 1 6.9 802761 2016 1 4 1 6 802981 2013 0 1 3 7.3 802981 2014 0 1 3 7.4 802981 2015 0 1 3 6.9 802981 2016 0 1 3 6 802981 2017 0 . 3 4.9 803099 2019 . 5 1 3.4 803201 2011 1 . 2 5 803201 2015 0 . 3 6.9 803295 2008 0 6 2 3.7 803295 2009 0 4 2 4.4 803356 2017 0 1 4 4.9 803536 2015 0 4 1 6.9 803567 2013 0 1 4 7.3 803567 2014 0 1 4 7.4 803567 2015 0 1 4 6.9 803567 2016 0 1 4 6 803567 2017 0 1 4 4.9 803567 2018 . . 4 3.8 803570 2014 0 5 1 7.4 803570 2019 . 4 1 3.4 803582 2014 0 2 2 7.4 803582 2017 0 . 3 4.9 803695 2017 0 . 2 4.9 803717 2008 1 4 2 3.7 803717 2011 0 3 2 5 803717 2012 0 4 3 5.8 803717 2013 0 2 3 7.3 803717 2014 0 2 3 7.4 803717 2015 0 2 3 6.9 803717 2016 0 2 3 6 803717 2017 0 1 4 4.9 803717 2018 0 1 4 3.8 803717 2019 . . 4 3.4 803747 2010 1 3 2 5 803747 2013 0 3 2 7.3 803747 2014 0 . 2 7.4 803747 2016 0 2 2 6 803747 2017 0 2 2 4.9 803775 2015 0 1 2 6.9 803839 2009 0 5 2 4.4 803839 2010 0 4 2 5 803839 2011 1 4 2 5 803854 2012 0 5 2 5.8 803854 2013 1 3 2 7.3 803854 2015 0 4 2 6.9 803854 2016 0 4 2 6 803854 2017 1 4 2 4.9 803906 2016 0 2 2 6 803906 2017 0 . 2 4.9 803924 2016 1 6 1 6 803952 2017 1 1 3 4.9 804019 2016 1 5 2 6 804031 2008 0 1 3 3.7 804031 2010 0 1 3 5 end label values self_percp_nor i257_7gl label def i257_7gl 1 "0%", modify label def i257_7gl 2 "Very Unlikely", modify label def i257_7gl 3 "Unlikely", modify label def i257_7gl 4 "Neutral", modify label def i257_7gl 5 "Likely", modify label def i257_7gl 6 "Very Likely", modify label def i257_7gl 7 "100%", modify label values age_cat agecat_2lb label def agecat_2lb 1 "26-39", modify label def agecat_2lb 2 "40-54", modify label def agecat_2lb 3 "55-59", modify label def agecat_2lb 4 "60-66", modify
