Do we always have to run a regression to estimate the treatment effect in a Differences in Differences Model? Yes or no.View keyboard shortcutsEditViewInsertFormatToolsTable12ptParagraph

* ** we ****** **** ** *** a ********** ** ******** *** ********* effect in a *********** in *********** ********** *** ** we **** *** ****** ****** ********** ** a *********** ** *********** ******** usually ***** trends as **** as ** **** ** follow **** trend *** conclusion ***** be **** ** ***** common ***** experienced ****** ****** *** ****** test ********* ** ***** assumption3 ** *** ** ***** ** **** ***** *** ******* ***** ******* ***** Fixed ******* **** ********** was *** basis for *********** *** ********* ********** **** ***** is *** ******** in traffic ******* fatalities *** ***** ** **** **** impact ** ********* ** **** *** ******** *** up having **** ********** to **** ****** The ************ ** not ****** ***** ***** *** ***** ********* ***** **** ** ** ********* ** *** DD ***** ** **** ***** *** ******* ***** with state *** **** Fixed ******* **** comparison *** *** ***** for *********** the ********* ********** **** ********* ***** **** ******** ********* more beer **** ******* **** tax cut hence *** ***** production ***** lead ** **** consumption ***** more ***** *********** *** ** ** use Fixed ******* ** *** ***** *** what ** the ********* *** ********* ************** ** *** **** ********* about ********* ****** of ******** ****** **** ***** time *** *** over ***** ********* since ***** ****** model ** **** **** ********* relationships between *********** as **** ** dependent variable ** * specified entity6 **** two ******* *** ************ a ** ********* PROC ********* PROC ******* **** *** **** ***** **** ********* FEs ** * model *** explain *** *** ******* (ie the ****** ** *** ********** ***** *** ******* ** FEs- it ** **** **** when ********* ****** ** ********* ***** ****** **** time *** *** over ********* there *** ****** ***** could come ** in ***** ******* ********** ** *** ******** with ****** ****** lagged dependent ********* *** there *** **** ** *** around *** ***** ********* ** ******** ** ** *** **** *** *** ************ of ***** the approach?- the ****** ***** ***** **** ** in fixed ******* estimation if ****** combined with ****** ****** ****** ********* ******** ** do **** **** **** *** lagged ********* variable **** to be ******* ** edited9 **** *** *** **** *** cons ** ***** * ****** *********** *********** they take **** account ********* variables *** every *********** **** ** ** *** relationship ******* *** simpleCons: **** *** ******** *********** ********** ****** restrictions *** put ** on **** it *** give *** ********* ************ to **** ************* ******* **** ********** What *** *** pros and cons ** using * Nonlinear (ie Probit ** ****** *********** ** well ** ****** ** differ ** *** *** **** ****** f *** Logit ***** utilizes ****** *** cumulative ************ ******** *** *** logistic distribution while probit ***** ******** ********** distribution ******** *** *** standard ****** distribution ** *** process of ******** * ******** they ******** any ****** ** **** ** ******* ** fall **** ** * *** allows ** ** transformed- **** ***** similar *********** Can **** ** used ** ****** ******** such ** ****************** Only ****** ****** ***** ** *********** ***** ********** *** *** non-constant error variances- * ****** ** ***** ***** *** only used ** ********** ******** ******* ****** ******* an ********* *********** *** *** ******* ********** ********** generates ********* ** *** ***** **** compare that ******* ** how OLS does itMaximum ********** *********** ** ** ******** **** *** ********* ******* **** ************ are *** dependent *** chance ** having ********** ** L=f(x1a)…f(a ** *** *** ** ********* as a *** **** ** ********** parameters ** ***** ****** ** ** *** *** *********** ***** ** are ****** ********** ********** with ********* = ** * B1*X- ** always *** ** ******* the model in *** ******** ********** ** *** ** assuming ** **** ******** ****** of * and ** *** ******** ***** Squares ***** can ** **** ** fitting *** ****** ********** ***** *** ******** B0 *** ** But ***** to *** *** *** ******** ********** ** ***** ******** ** **** something ** observe12 Compare time ****** *** *************** dataWith *** time series **** *** observations are ** ****** ******* given at ******** **** ******* ***** on ***** sectional **** it consists of *** observations ***** include many ******** in **** time ****** ** interval *** ***** of time series **** ** ** same ******** ******** **** **** while ***** sectional **** looks ** ******* variables in *** point in time13 Name two methods *** ********* autocorrelation- *** ****** ****** ***** Breusch-Godfrey ** **** for ***************** **** *** ******* *** ** *** *** ******* *** ********** ****************** can ******* ***** ******** in *** **** **** helps reduce the **** ************* Name *** ******* the main concern **** ******* **** ************* ************ *** ********** ** be *** *********** ***** ***** the ********* ** *********** **** ***** ***** *** ********************* ** ***** **** *** results gotten could indicate ************ between ********* besides it ***** *** ***** hence results ***** *** ******* at *** be unreliable ******* ** ***** ************** How ** ** test *** **************** ************ ** the **** series ** * given **** *** ** tested ***** parametric ***** ** utilizing statistical test developed *** **** ******* to ****** ********** ***** ** stationarity *** ************* **** ** ****** method use ** **** **** ************ *** ********* are steps *** *** ***** ******* I Test *** * unit ***** ∆yᵢ * ************ * **** ** **** *** * **** root **** ****** ∆yᵢ * **** + ************ * **** *** **** *** * unit root **** ***** *** ************* **** ************* = **** + ****** + ************ * ****** Name *** ****** ** ********** for ****************** **** ****