Hey guys
I'm working on a project studying the effect of sms-nudges on student absence in (cumpulsory) primary and lower secondary school in Denmark.
I estimate the effects with a fixed effects regression design with fixed effects for school and school year. It is my identifying assumption that the two groups (the nudged and non-nudged) are balanced, and i wish to demonstrate this in my project through a socioeconomic reference variable. This variable shows the statistically expected mean of grades based on a variety of backgroundvariables of the students attending a specific school. To sum up - i wish to demonstrate that my control- and treatmentgroup is balanced on the socioeconomic reference variable.
I was advised by a professor to run a fixed effects regression with the sms-nudge variable as the independent and the socioeconomic reference as dependent to investigate if there is a change in socioeconomic ref. after the nudge. And of course i hope this is not the case, as this means the two groups are balanced. However, i'm not quite sure that i interpret the output from the regression correctly.
The specific command: reghdfe socio_ref smsnudge, absorb(schoolyear inst_nr) cluster(inst_nr)
Question: does this (see output) show that the variation in socioeconomic reference between schools is controlled for with my fixed effects and therefore that the control- and treatment group is balanced (my identifying assumption is confirmed)?
... or is the output of no use for what i wish to demostrate?
I hope some of you can help me with this!
I'm working on a project studying the effect of sms-nudges on student absence in (cumpulsory) primary and lower secondary school in Denmark.
I estimate the effects with a fixed effects regression design with fixed effects for school and school year. It is my identifying assumption that the two groups (the nudged and non-nudged) are balanced, and i wish to demonstrate this in my project through a socioeconomic reference variable. This variable shows the statistically expected mean of grades based on a variety of backgroundvariables of the students attending a specific school. To sum up - i wish to demonstrate that my control- and treatmentgroup is balanced on the socioeconomic reference variable.
I was advised by a professor to run a fixed effects regression with the sms-nudge variable as the independent and the socioeconomic reference as dependent to investigate if there is a change in socioeconomic ref. after the nudge. And of course i hope this is not the case, as this means the two groups are balanced. However, i'm not quite sure that i interpret the output from the regression correctly.
The specific command: reghdfe socio_ref smsnudge, absorb(schoolyear inst_nr) cluster(inst_nr)
Question: does this (see output) show that the variation in socioeconomic reference between schools is controlled for with my fixed effects and therefore that the control- and treatment group is balanced (my identifying assumption is confirmed)?
... or is the output of no use for what i wish to demostrate?
I hope some of you can help me with this!
Comment