看看代码和回归结果，学习下

你好，我又学了个别的方法。这次变成了政策前系数远大于0，回归后接近0.和别人完全相反了，这是为什么呢

给出你的结果再说

代码是这样的：set more off
xtset city year
gen distance = year - policy_year
replace distance = -4 if distance < -4
*replace distance = 5 if distance  > 5
forvalues i=1/4 {
gen d_`i'  = 0
replace d_`i'  = 1 if treat== 1 & distance== -`i'
}
forvalues i=1/5 {
gen d`i'  = 0
replace d`i'  = 1 if treat== 1 & distance== `i'
}
gen current  = 0
replace current = 1 if treat== 1 & distance== 0
xtreg y d_4 d_3 d_2 d_1 d1 d2 d3 d4 d5 i.year x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 , fe r
coefplot,keep( d_4 d_3 d_2 d_1 d1 d2 d3 d4 d5 ) levels(90) vertical lcolor(black) mcolor(black) msymbol(circle_hollow) ytitle(回归系数, size(small))  ylabel(, labsize(small) angle(horizontal) nogrid) yline(0, lwidth(vthin)lpattern(solid) lcolor(black)) xtitle(政策实施相对时间, size(small)) xlabel(,labsize(small))  graphregion(fcolor(white) lcolor(white) ifcolor(white) ilcolor(white)) ciopts(recast(rcap)) xline(10.5, lwidth(vthin) lpattern(solid)lcolor(black))

set more off
xtset city year
gen distance = year - policy_year
replace distance = -4 if distance < -4
*replace distance = 5 if distance  > 5
forvalues i=1/4 {
gen d_`i'  = 0
replace d_`i'  = 1 if treat== 1 & distance== -`i'
}
forvalues i=1/5 {
gen d`i'  = 0
replace d`i'  = 1 if treat== 1 & distance== `i'
}
gen current  = 0
replace current = 1 if treat== 1 & distance== 0
xtreg y d_4 d_3 d_2 d_1 d1 d2 d3 d4 d5 i.year x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 , fe r
coefplot,keep( d_4 d_3 d_2 d_1 d1 d2 d3 d4 d5 ) levels(90) vertical lcolor(black) mcolor(black) msymbol(circle_hollow) ytitle(回归系数, size(small))  ylabel(, labsize(small) angle(horizontal) nogrid) yline(0, lwidth(vthin)lpattern(solid) lcolor(black)) xtitle(政策实施相对时间, size(small)) xlabel(,labsize(small))  graphregion(fcolor(white) lcolor(white) ifcolor(white) ilcolor(white)) ciopts(recast(rcap)) xline(10.5, lwidth(vthin) lpattern(solid)lcolor(black))