***** STATA DO FILE MASTER THESIS **** Dataset * Renaming variables rename var1 ID_Activism rename var2 Company rename var3 Techcompany rename var4 Log_Employees rename var5 Log_TotalAssets rename var6 Date rename var7 Year rename var8 CAR rename var9 Topics rename var10 Issues rename var11 Cases rename var12 ActivismPubPolicy rename var13 ActivismType rename var14 ActivismControversy rename var15 TopicNew rename var16 IssueNew rename var17 CaseNew * Categorical variables encode Company, gen (Companies) drop Company tostring Year, replace encode Year, gen (Years) drop Year encode Topics, gen (Topic) drop Topics encode Issues, gen (Issue) drop Issues encode Cases, gen (Case) drop Cases * Dummy variables tabulate Techcompany, gen (Tech_company) drop Tech_company1 Techcompany rename Tech_company2 Tech_company label variable Tech_company "Tech company=1, No=0" tabulate ActivismPubPolicy, gen (Activism_PubPolicy) drop Activism_PubPolicy1 ActivismPubPolicy rename Activism_PubPolicy2 Activism_PubPolicy label variable Activism_PubPolicy "Reaction PubPolicy=1, No=0" tabulate ActivismType, gen (Activism_Type) drop Activism_Type2 ActivismType rename Activism_Type1 Activism_Type label variable Activism_Type "CEO=1, No=0" tabulate ActivismControversy, gen (Activism_Controversy) drop Activism_Controversy1 ActivismControversy rename Activism_Controversy2 Activism_Controversy label variable Activism_Controversy "High=1, Moderate=0" tabulate TopicNew, gen (Topic_new) drop Topic_new1 TopicNew rename Topic_new2 Topic_new label variable Topic_new "Topic new=1, No=0" tabulate IssueNew, gen (Issue_new) drop Issue_new1 IssueNew rename Issue_new2 Issue_new label variable Issue_new "Issue new=1, No=0" tabulate CaseNew, gen (Case_new) drop Case_new1 CaseNew rename Case_new2 Case_new label variable Case_new "Case new=1, No=0" order ID_Activism Companies Tech_company Log_Employees Log_TotalAssets Date Years CAR Activism_Type Activism_PubPolicy Activism_Controversy Topic Topic_new Issue Issue_new Case Case_new *Check variables sum ID_Activism Date CAR Companies Years Topic Issue Case Tech_company Activism_PubPolicy Activism_Type Activism_Controversy Topic_new Issue_new Case_new Log_Employees Log_TotalAssets ** 111 observations of Activism_Controversy, 11 missing values. Considered ok. ** 51 observations of Topic_new, Issue_new, Case_new, need to be addressed. *** Missing values ** Test regression reg CAR Companies Tech_company Years Activism_Type Activism_Controversy Activism_PubPolicy Topic Topic_new Issue Issue_new Case Case_new Log_Employees Log_TotalAssets reg CAR Topic Topic_new Issue Issue_new Case Case_new * Only 51 variables, not significant model/IVs ** Assume missing values = 0 (not new) replace Topic_new = 0 if (Topic_new >0 .) replace Issue_new = 0 if (Issue_new >0 .) replace Case_new = 0 if (Case_new >0 .) sum Topic_new Issue_new Case_new * Gives 122 observations on each Newness variable ** Correlation matrix pwcorr CAR Companies Tech_company Log_Employees Log_TotalAssets Years Activism_Type Activism_PubPolicy Activism_Controversy Topic Topic_new Issue Issue_new Case Case_new, sig star(.05) * We see that the newness variables are omitted due to missing values. These are therefore excluded from further analysis. * Further, we find high correlations for Years, and the variable is therefore excluded from further analysis. *** Regression ** A stepwise backwards regression is performed. * Step 1: reg CAR Companies Tech_company Log_Employees Log_TotalAssets Activism_Type Activism_PubPolicy Activism_Controversy Topic Issue Case * Step 2: reg CAR Tech_company Log_Employees Log_TotalAssets Activism_Type Activism_PubPolicy Activism_Controversy Topic Issue Case * Step 3: reg CAR Tech_company Log_Employees Activism_Type Activism_PubPolicy Activism_Controversy Topic Issue Case * Step 4: reg CAR Tech_company Log_Employees Activism_Type Activism_PubPolicy Activism_Controversy Issue Case * Step 5: reg CAR Tech_company Log_Employees Activism_Type Activism_PubPolicy Activism_Controversy Issue * Step 6: reg CAR Tech_company Activism_Type Activism_PubPolicy Activism_Controversy Issue * Step 5: reg CAR Activism_Type Activism_PubPolicy Activism_Controversy Issue *Step 6: reg CAR Activism_Type Activism_Controversy Issue * Further investigation of Issue - specify indicators for each category of the Issue variable. reg CAR Activism_Type Activism_Controversy i.Issue * Further investigation of Issue versus controversy - VIF reg CAR Activism_Type Activism_Controversy i.Issue vif * Exploring other categorical variables reg CAR Activism_Type Activism_Controversy i.Topic reg CAR Activism_Type Activism_Controversy i.Case *** Exploring inetraction effects ** Potential interaction effects reg CAR Activism_Controversy##Activism_Type Activism_Controversy##Activism_PubPolicy Activism_Controversy##Tech_company Activism_Controversy##Topic reg CAR Activism_Controversy##Activism_Type reg CAR Activism_Controversy##Activism_PubPolicy reg CAR Activism_Controversy##Tech_company reg CAR Activism_Controversy##Topic reg CAR Activism_Controversy##Issue reg CAR Activism_Controversy##Case reg CAR Activism_Type##Activism_PubPolicy reg CAR Activism_Type##Tech_company reg CAR Activism_Type##Topic reg CAR Activism_Type##Issue reg CAR Activism_Type##Case reg CAR Tech_company##Activism_PubPolicy reg CAR Tech_company##Topic reg CAR Tech_company##Issue reg CAR Tech_company##Case