{smcl} {* 30aug2011}{...} {cmd:help rscore} {hline} {title:Title} {p2colset 5 18 20 2}{...} {p2col:{hi:rscore}{hline 1}}Estimation of responsiveness scores{p2colreset}{...} {title:Syntax} {p 8 17 2} {cmd:rscore} {it:outcome} [{it:varlist}] {ifin} {weight}{cmd:,} {cmd:model}{cmd:(}{it:{help rscore##modeltype:modeltype}}{cmd:)} {cmd:rs_name}{cmd:(}{it:stub}{cmd:)} [{cmd:factors}{cmd:(}{it:varlist_f}{cmd:)} {cmd:xlist}{cmd:(}{it:varlist_c}{cmd:)} {cmd:graph}{cmd:(}{it:#}{cmd:)} {cmd:radar}{cmd:(}{it:numlist}{cmd:)} {cmd:id_string}{cmd:(}{it:varname}{cmd:)} {cmd:vce}{cmd:(}{it:vcetype}{cmd:)} {cmd:save_graph1}{cmd:(}{it:filename}{cmd:)} {cmd:save_graph2}{cmd:(}{it:filename}{cmd:)}] {pstd}{cmd:fweight}s, {cmd:pweight}s, {cmd:iweight}s are allowed; see {help weight}. {title:Description} {pstd} {cmd:rscore} computes unit-specific responsiveness scores using an iterated Random-Coefficient-Regression (RCR). The basic econometrics of this model can be found in Wooldridge (2002, pp. 638-642). The model estimated by {cmd:rscore} considers a regression of a response variable y, i.e. ({it:outcome}), on a series of factors (or regressors) x, i.e. {it:varlist}, by assuming a different reaction (or "responsiveness") of each unit to each factor contained in x. {cmd:rscore} allows for: (i) ranking units according to the level of the responsiveness score obtained; (ii) detecting factors that are more influential in driving unit performance; (iii) studying, more in general, the distribution (heterogeneity) of factor responsiveness scores across units. {title:Options} {phang} {cmd:model}{cmd:(}{it:{help rscore##modeltype:modeltype}}{cmd:)} specifies the model to be estimated, where {it:modeltype} must be one of the following models: "ols", "fe", "re". It is always required to specify one model. {phang} {cmd:rs_name}{cmd:(}{it:stub}{cmd:)} specifes the name the user wants to give to the responsiveness scores variables generate by rscore. RS variables are thus named as: {it:stub}1, {it:stub}2, {it:stub}3, ... and so forth. {phang} {cmd:factors}{cmd:(}{it:varlist_f}{cmd:)} specifies that factor variables have to be included among the regressors. It is optional for both models. {phang} {cmd:xlist}{cmd:(}{it:varlist_c}{cmd:)} specifies that control variables (which are not factors) have to be included among the regressors. It is optional for both models. {phang} {cmd:graph}{cmd:(}{it:#}{cmd:)} provides a combined graph of the densities of the responsiveness scores. The number # defines the width of the graph's x-axis. The user can set a proper # for providing a good rendering of the graph. {phang} {cmd:radar}{cmd:(}{it:numlist}{cmd:)} provides a radar plot of the responsiveness scores for the units specifed in numlist. Notice that, in order to run this option, the user must specify the {cmd:id_string}{cmd:(}{it:varname}{cmd:)} option. {phang} {cmd:id_string}{cmd:(}{it:varname}{cmd:)} requests to specify a string variable as identifer of each observation. This is compulsory if the user wishes to provide a radar plot of the responsiveness scores. {phang} {cmd:vce}{cmd:(}{it:vcetype}{cmd:)}: allows to choose {it:vcetype} as either {it:robust}, or {it:cluster clustvar}. {phang} {cmd:save_graph1}{cmd:(}{it:filename}{cmd:)} allows to save the graph generate by the option {cmd:graph}{cmd:(}{it:#}{cmd:)} in the user specifed filename. {phang} {cmd:save_graph2}{cmd:(}{it:filename}{cmd:)} allows to save the graph generate by the option {cmd:radar}{cmd:(}{it:numlist}{cmd:)} in the user specifed filename. {marker modeltype}{...} {synopthdr:modeltype_options} {synoptline} {syntab:Model} {p2coldent : {opt ols}}regression estimated by ordinary least squares (OLS){p_end} {p2coldent : {opt fe}}panel data fixed-effect regression (FE){p_end} {p2coldent : {opt re}}panel data random-effect regression (RE){p_end} {synoptline} {pstd} {cmd:rscore} returns goodness-of-fit statistics - i.e. R-squared - for each estimated factor regression, by storing them in the following scalars: {bf:e(R1)}, ..., {bf:e(RQ)}. Further, {cmd:rscore} provides the average R-squared - i.e the overall goodness-of-fit of the model - stored in the scalar {bf:e(R)}. {pstd} {cmd:rscore} creates a number of variables: {pmore} {inp:{it:stub}j} is the responsiveness scores variable related to the {it:j}-th variable of {it:varlist}. They are as many as the variables considered in {it:varlist}. {title:Remarks} {pstd} Please, before running this program, remember to have the most recent up-to-date version installed. {title:Examples} {inp:. rscore y x1 x2 x3 , rs_name(RS) model(ols) factor(f1 f2)} {inp:. rscore y x1 x2 x3 , rs_name(RS) model(fe) factor(f1 f2) xlist(x4 x5)} {title:Reference} {phang} Wooldridge, J. M. 2002. {it: Econometric Analysis of Cross Section and Panel Data}. The MIT Press, Cambridge. {p_end} {title:Author} {phang}Giovanni Cerulli{p_end} {phang}IRCrES-CNR{p_end} {phang}Research Institute for Sustainable Economic Growth, National Research Council of Italy{p_end} {phang}E-mail: {browse "mailto:giovanni.cerulli@ircres.cnr.it":giovanni.cerulli@ircres.cnr.it}{p_end} {title:Also see} {psee} Online: {helpb ivregress} {p_end}