*These data have already been transformed from the initial TRIP data to evaluate the claims in Maliniak et al. 2013 *Citation data comes from the Web of knowledge. These values were gathered by an automated script in March 2013 and linked to *the articles in the TRIP database by the unique combination of values formed by *the title, journal of publication, issue number, and volume number+ The number of *citations for each article reflects citations from all articles catalogued in the WOK’s *Social Science Citation Index Expanded, not just those journals from which we *draw our sample. See Maliniak et al. 2013 for more information. use "replication_data.dta", clear rename id cited_pub_id merge 1:1 cited_pub_id using "citation_counts_age_wide.dta" rename cited_pub_id id drop _merge merge 1:1 id using "processed_gender_trip.dta", keepusing(authscore*) force drop _merge save Maliniak_et_al_replication_data.dta, replace drop if C__Year>2007 *table1 estimates clear estpost tabstat sscie if in_analysis==1 & C__Year > 1980 & C__Year < 1990, by(gender_comp) statistics(mean sd median) listwise eststo a estpost tabstat sscie if in_analysis==1 & C__Year > 1990 & C__Year < 2000, by(gender_comp) statistics(mean sd median) listwise eststo b estpost tabstat sscie if in_analysis==1 & C__Year > 2000 & C__Year < 2007, by(gender_comp) statistics(mean sd median) listwise eststo c esttab a b c using table1.rtf, replace cells("mean(fmt(2)) sd(fmt(2)) p50(fmt(0))") nostar label nonotes mtitles("1980s" "1990s" "2000s") nonumbers title("Table 1: Citations by gender and decade.") *Table 2 *simple baseline nbreg sscie_count all_female coed y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo gender *add age nbreg sscie_count all_female coed article_age article_age_sq y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo age *Career Model nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo career *Epistemology nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo epist *Material nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist material y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo material *Ideational nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist material idea y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo idea *Paradigm nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist material idea para_atheo-para_real y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo paradigm *Issue nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist material idea para_atheo-para_real y1980-y2007 issue_american-issue_usfp if in_analysis == 1 & C__Year < 2007, robust eststo issue *Methodology nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist material idea para_atheo-para_real y1980-y2007 issue_american-issue_usfp meth_quant-meth_count if in_analysis == 1 & C__Year < 2007, robust eststo method *kitchen_sink nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist material idea para_atheo-para_real y1980-y2007 issue_american-issue_usfp meth_quant-meth_count AJPS APSR BJPS EJIR IO IS ISQ JCR JOP JPR SS WP if in_analysis == 1 & C__Year < 2007, robust eststo kitchen_sink esttab gender age career epist material idea paradigm issue method kitchen_sink using table6.rtf, replace label nodepvars nogaps nonumbers mtitles mtitles("Gender" "Time" "Career" "Epistemology" "Material" "Ideational" "Paradigm" "Issue Area" "Methodology" "Kitchen Sink") se(%8.3g) b(%8.3g) starlevels(* 0.10 ** 0.05 *** 0.01) drop(y1* y2* o*) stat(N ll aic r2_p, labels("N" "Log Lik." "AIC" "R2") fmt(%10.7g %10.7g %10.7g %7.2g)) *using a model to predict female citations nbreg sscie_count article_age article_age_sq tenured tenured_female coauthored R1 y1980-y2007 if in_analysis == 1 & C__Year < 2007&all_male==1, robust predict m1, n predict m1_ci, stdp gen m1_upper = m1+(m1_ci*1.96) gen m1_lower = m1-m1_ci*(1.96) corr m1 sscie_count if all_female==1 gen diff_citation_count= sscie_count-m1 bysort gender_comp: sum diff_citation_count m1 sscie_count if in_analysis==1, detail ttest m1=sscie_count if all_male==1&in_analysis==1 ttest m1=sscie_count if all_female==1&in_analysis==1 ttest m1=sscie_count if coed==1&in_analysis==1 ttest diff_citation_count, by( all_female) label variable m1 "Predicted" label variable sscie_count "Actual" label variable diff_ "Diff." estimates clear estpost tabstat m1 sscie diff_ if in_analysis==1 & C__Year < 2007, by(gender_comp) statistics(mean) columns(statistics) listwise esttab using table7.rtf, replace main(mean) nostar nogaps unstack label nonotes nomtitles nonumbers title("Table 7: Predicted vs. Actual Citation Counts.") *Models with weighting, i.e. parametric matching mlogit gender article_age article_age_sq tenured coauthored R1 positivist material idea para_atheo-para_real issue_american-issue_usfp meth_quant-meth_count APSR BJPS EJIR IO IS ISQ JCR JOP JPR SS WP if in_analysis == 1 & C__Year < 2007, robust predict ps1 ps2 ps3 gen dose_w=1/ps1 if gender==1 replace dose_w=1/ps2 if gender==2 replace dose_w=1/ps3 if gender==3 nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 y1980-y2007 if in_analysis == 1 & C__Year < 2007, robust eststo m1 nbreg sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 y1980-y2007 if in_analysis == 1 & C__Year < 2007 [pw= dose_w], robust eststo m2 esttab m1 m2 using table7a.rtf, replace label nogaps nonumbers mtitles mtitles("Unweighted" "Weighted") se(%8.3g) b(%8.3g) starlevels(* 0.10 ** 0.05 *** 0.01) drop(y1* y2* o*) stat(N ll aic r2_p, labels("N" "Log Lik." "AIC" "R2") fmt(%10.7g %10.7g %10.7g %7.2g)) title("Table 7a: Matching Analysis Confirms the Gap Exists.") */ ************************ * Is it getting better? ************************ *Interaction graph quietly summarize article_age, detail local min=5 local max=34 local cen25=r(p25) local cen50=r(p50) local cen75=r(p75) local numparams=20 local inc=(`max'-`min')/(`numparams'-1) local iter=0 matrix foo2 = 0,0,0,0 local order=2 while `iter'<`numparams' { gen article_agea=article_age-`min'-(`inc'*`iter') summarize article_agea gen article_ageaXall_female=article_agea*all_female gen article_age_sqa= article_agea^2 gen article_age_sqaXall_female=article_age_sqa*all_female display `min' display `inc' display `iter' nbreg sscie_count article_agea all_female article_ageaXall_female tenured R1 y1980-y2008 if C__Year <= 2007 & C__Year >= 1980 & in_analysis == 1 matrix betas=e(b) scalar x1coef=betas[1,`order'] matrix ses=e(V) scalar x1se=sqrt(ses[`order',`order']) local obs=e(N) scalar ci95=invttail(`obs', 0.05) local xval = `min'+(`inc'*`iter') matrix foo = x1coef-ci95*x1se, x1coef, x1coef+ci95*x1se, `xval' matrix foo2 = foo2 \ foo drop article_agea* drop article_age_sqa* local iter=`iter'+1 } matrix points=foo2[2..(`numparams'+1),1..4] svmat points twoway (rconnected points1 points3 points4, scheme(lean1) scale(.8) lpattern(-) msymbol(i) lcolor(black)) (histogram article_age if article_age < 35 & article_age >=5 & in_analysis == 1, width(1) yaxis(2) blcolor(gray) bfcolor(none)) (connected points2 points4, msymbol(i) lcolor("red")), ylabel(, labsize(*medium)) ytitle(Coeff. on All Female and 95% CIs) ytitle(Histogram of Article Age, axis(2)) yline(0, lwidth(medthick)) xtitle(Article Age) xlabel(5(5)34, labsize(small)) title("Coeff. on All Female at Different Levels of Article Age") legend(off) *Only for Mac * graph export "figure5.pdf", replace drop points* clear matrix gen article_ageXall_female = all_female*article_age nbreg sscie_count article_age all_female article_ageXall_female tenured R1 y1980-y2008 if C__Year <= 2007 & C__Year >= 1980 & in_analysis == 1 eststo getting_better esttab getting_better using getting_better.rtf, replace nogaps nonumbers mtitles mtitles("Full Sample") se(%8.3g) b(%8.3g) starlevels(* 0.10 ** 0.05 *** 0.01) drop(y1* y2* o*) stat(N ll aic r2_p, labels("N" "Log Lik." "AIC" "R2")) sfmt(%8.2g) title("Table 9: Subsample analysis of citation count at six years.") label *full sample nbreg cum_citations6 all_female coed coauthored tenured tenured_female R1 y1980-y2007 if C__Year >=1980 & C__Year <= 2007 & in_analysis == 1 eststo model_all *1980s sub sample nbreg cum_citations6 all_female coed coauthored tenured tenured_female R1 y1980-y1989 if C__Year >=1980 & C__Year <= 1989 & in_analysis == 1, robust eststo model_80 *1990s sub sample nbreg cum_citations6 all_female coed coauthored tenured tenured_female R1 y1990-y1999 if C__Year >=1990 & C__Year <= 1999 & in_analysis == 1, robust eststo model_90 *2000s sub sample nbreg cum_citations6 all_female coed coauthored tenured tenured_female R1 y2000-y2007 if C__Year >=2000 & C__Year <= 2008 & in_analysis == 1, robust eststo model_00 esttab model_all model_80 model_90 model_00 using table9.rtf, replace nogaps nonumbers mtitles mtitles("Full Sample" "1980s" "1990s" "2000s") se(%8.3g) b(%8.3g) starlevels(* 0.10 ** 0.05 *** 0.01) drop(y1* y2* o*) stat(N ll aic r2_p, labels("N" "Log Lik." "AIC" "R2")) sfmt(%8.2g) title("Table 9: Subsample analysis of citation count at six years.") label gen auth_six_years = . replace auth_six_years=authscore_1986 if C__Year==1980 replace auth_six_years=authscore_1987 if C__Year==1981 replace auth_six_years=authscore_1988 if C__Year==1982 replace auth_six_years=authscore_1989 if C__Year==1983 replace auth_six_years=authscore_1990 if C__Year==1984 replace auth_six_years=authscore_1991 if C__Year==1985 replace auth_six_years=authscore_1992 if C__Year==1986 replace auth_six_years=authscore_1993 if C__Year==1987 replace auth_six_years=authscore_1994 if C__Year==1988 replace auth_six_years=authscore_1995 if C__Year==1989 replace auth_six_years=authscore_1996 if C__Year==1990 replace auth_six_years=authscore_1997 if C__Year==1991 replace auth_six_years=authscore_1998 if C__Year==1992 replace auth_six_years=authscore_1999 if C__Year==1993 replace auth_six_years=authscore_2000 if C__Year==1994 replace auth_six_years=authscore_2001 if C__Year==1995 replace auth_six_years=authscore_2002 if C__Year==1996 replace auth_six_years=authscore_2003 if C__Year==1997 replace auth_six_years=authscore_2004 if C__Year==1998 replace auth_six_years=authscore_2005 if C__Year==1999 replace auth_six_years=authscore_2006 if C__Year==2000 *full sample reg auth_six_years all_female coed coauthored tenured tenured_female R1 y1980-y2007 if C__Year >=1980 & C__Year < 2007 & in_analysis == 1 eststo model_all *1980s sub sample reg auth_six_years all_female coed coauthored tenured tenured_female R1 y1980-y1989 if C__Year >=1980 & C__Year <= 1989 & in_analysis == 1, robust eststo model_80 *1990s sub sample reg auth_six_years all_female coed coauthored tenured tenured_female R1 y1990-y1999 if C__Year >=1990 & C__Year <= 1999 & in_analysis == 1, robust eststo model_90 *2000s sub sample reg auth_six_years all_female coed coauthored tenured tenured_female R1 y2000-y2007 if C__Year >=2000 & C__Year <= 2008 & in_analysis == 1, robust eststo model_00 esttab model_all model_80 model_90 model_00 using table10.rtf, replace nogaps nonumbers mtitles mtitles("Full Sample" "1980s" "1990s" "2000s") se(%8.3g) b(%8.3g) starlevels(* 0.10 ** 0.05 *** 0.01) drop(y1* y2* o*) stat(N ll aic r2, labels("N" "Log Lik." "AIC" "R2")) sfmt(%5.2g) title("Table 10: Subsample analysis of authority score at six years.") label label variable sscie_count "Citation Count" *descriptive stats table estimates clear estpost summarize sscie_count all_female coed article_age article_age_sq tenured tenured_female coauthored R1 positivist material idea para_atheo-para_real C__Year issue_american-issue_usfp meth_quant-meth_count AJPS APSR BJPS EJIR IO IS ISQ JCR JOP JPR SS WP if in_analysis==1 & C__Year < 2007,listwise esttab using table5.rtf, replace cells("mean(fmt(2)) sd(fmt(2)) min max") nostar nogaps label nonotes nonumbers nomtitles title("Table 5: Descriptive Statistics") estimates clear estpost tabulate Issue_Area gender_comp esttab using table2.rtf, replace cell(colpct(fmt(2))) unstack noobs label varlabels(`e(labels)') nonumbers nodepvars nonotes nomtitles title("Table2: Gender of Authors by Issue Area.") *table 3 estimates clear estpost tabulate Paradigm gender_comp esttab using table3.rtf, replace cell(colpct(fmt(2))) unstack noobs label varlabels(`e(labels)') nonumbers nodepvars nonotes nomtitles title("Table2: Gender of Authors by Paradigm.")