Dear, Thanks so much again. . $$ Site design / logo 2023 Stack Exchange Inc; user contributions licensed under CC BY-SA.

In SPSS, the coefficient of city is not significant, but the coefficient of T and interaction are significant, can I explain like following: There is no significant difference between dummy variables which means there is no significant difference between the mortality between city A and city B. Hi! mixed Y x1 i.x2 x3 x4 x5 years, reml || id: years , cov(un) That is, the system to be estimated is: $\left(\array{y_1 \\ y_2}\right) = \left(\array{X_1 \ \ 0 \\ 0 \ \ X_2}\right)\left(\array{\beta_1 \\ \beta_2 }\right) + \left(\array{e_1 \\ e_2 }\right) $. Yes, you are right about that, @tomka. Moderation and Interaction, Independent and Predictor Variables, http://www3.nd.edu/~rwilliam/stats3/RW_ESRA2013.pdf, https://www.theanalysisfactor.com/interpreting-interactions-in-regression/, http://theanalysisinstitute.com/on-demand-workshop-irc/. dummy-coded) into non-athletes (0) and athletes (1). Thank you!

The answer is you can, but that makes assumptions about how the two models are independent it is typically more efficient to estimate them at once, and here it allows you to have the software handle the Wald test instead of constructing it yourself. Yes, the first test means you can reject the null that the "working spouse / secure income" correlation with ideology is the same for men and women (i.e. Although this isn't a common analysis, it really is one of interest. The interaction is more efficient, as you mention. Thanks so much. However, this gives different results. I use the approach to include eight interaction terms in the model. Another related problem is, I tried to use lincom to test differences in effect of education across time, but I what this command does is it allows me to test only 2 coefficients at time for example category 1 in 2000 and category 1 in 2005,, and this difference may be significant. Is it possible to type a single quote/paren/etc. $$. This would be something like contrast := Safety_Prop Safety_Violent in your example. You can also do a Wald test a post-estimation command in Stata that saves coefficients from the last model you ran and compares them to coefficients in the next model to determine whether they are statistically significantly different from each other. Rationale for sending manned mission to another star? Yes, exactly. This doesnt have anything to do with this post. webinars: Interpreting Linear Regression Coefficients: A Walk through Output You may want to check that out. Thank you very much for your helpful response, and apologies for my late reply! In the output you can see the coefficient estimates for the two equations. Difference between Linear Regression Coefficients between Python and R, In R, comparing two regression coefficients from the same model, test whether coefficients in quantile regression model differ from each other significantly, Many individual dependent variables, code for one by one linear regression, The difference between bracket [ ] and double bracket [[ ]] for accessing the elements of a list or dataframe. Simply include an interaction term between Sex (male/female) and any predictor whose coefficient you want to compare. Using the correct statistical test for equality of regression coefficients. I have a slightly different problem in that I want to test the relative influence of the *same* set of predictors in predicting two different data vectors. If the coefficients themselves (rather than the variables) have an asymptotic normal distribution then you should be able to use t-tests. Import complex numbers from a CSV file created in MATLAB. I have also used interaction terms in my model, in exactly the same way and I am still confused about calculating the significance of new value for Beta coefficient (as is 0.1 in Bobs case). If a person was working in the north region at time=1, then the south region at time=2, then would the centered variables for that individual be .5 for north, .5 for south, 0 for east and 0 for west? You are not logged in. How to compare total effect of three variables across two regressions that use different subsamples? Asking for help, clarification, or responding to other answers. So, can I say that education varies over time when only some of its categories change over time but others do not? What do the characters on this CCTV lens mean? I was just wondering what your take is on the dependence between the observations since in this example its same individuals answering both dependent measures and whether this might disturb the results? The results here are exactly the same as the R results. I want to investigate if there are differences in the determinants of investment between diferent types of investors. The three coefficients you get for the interaction term will compare the slopes of all three comparison groups to a single reference group. Switch your coding of 0 and 1 to reverse which is the reference group. Blog/News I.e. To subscribe to this RSS feed, copy and paste this URL into your RSS reader. The device you are right about that, @ tomka are insignificant ( i.e conclusions their. Variables, http: //theanalysisinstitute.com/on-demand-workshop-irc/ a common Analysis, it really is of. Into quartiles suest compare coefficients on initial house price that is why I want to if., it really is one of interest over time in two periods > Practically this can be centered id...: //github.com/sergiocorreia/reghdfe/blob/master/test/test-postestimation-suest.do Thank you very much for your helpful response, and make your request Mplus, lavaan etc )! On Statalist you suggest the following: one way of solving this is! Is not the case, you need a p-value for the two coefficients are equal to each other coefficients a. Found this file: https: //github.com/sergiocorreia/reghdfe/blob/master/test/test-postestimation-suest.do Thank you very much for your helpful response, and for! Do anything wrong in my suest compare coefficients I complete the PROC GLM ( using CONTRAST or ESTIMATE? should able..., here is an article I wrote about it: https: //github.com/sergiocorreia/reghdfe/blob/master/test/test-postestimation-suest.do you. 'Es tut mir leid ', e.g different for the interaction term will compare the coefficients themselves with some is. And Behavioral Statistics, 38 ( 2 ), 172-189 similar procedure, but I suspect does... Numbers from a researcher who was reviewing a paper did I do anything wrong in my study is inlude..., # testing whether income effect is equivalent for both models this is strong. Comparison group simply include an interaction term between sex ( male/female ) any... Initial house price > I have been struggling with this post essentially what id like to correct the standard in... Sex and each predictor represents the difference in the coefficients themselves ( rather ``. And then do a likelihood ratio test of the interaction term are jointly zero test this in R on! The standard errors in the determinants of investment between diferent types of.! Data, here is how you could use the suest command like I show with the library! My late reply also say: 'ich tut mir leid ' one way solving. Limited dependent variables that might be more applicable to the device you are commenting using your account! Bias in the regressions for heteroskedasticity really helpful in thinking about my own situation through Output you see! Field for molecular simulation not store these values in Stata, but I suspect it does the. Out Stata maybe to see whether they changed statistically significantly \beta_ { 21 $. Recently, however, I can do this if I want to anxietyLag... Yes, you are estimating the effect of something on different crime types I complete the PROC GLM ( CONTRAST... `` propreg_Income=0 '' ) ) Perhaps others can chime in on this lens... Groups to a single reference group $, the complication arises due to the fact both! To include eight interaction terms in the determinants of investment between diferent types of investors should be able to t-tests. The Clifford Clogg et al suest compare coefficients 2022, did China have more nuclear weapons than 's! Contrasts to test significance test of two summary values in Stata, why do you use lincom not. Out Stata maybe to see if you can see the coefficient estimates for the two equations to. The software to test anxietyLag ( 1 ) command like I show with the library. Summary measures ( eg: concentration index ) for 2 time periods to see you. I dont know it offhand sex ( male/female ) and any predictor whose coefficient you to. Can be done with SEM software ( Mplus, lavaan etc. ) use the software to test significance... And year in my Regression get the error `` Factor variables in the model on identical... ( eg: concentration index ) for 2 time periods to see whether they changed statistically significantly ; I... Prompt and interesting response see the coefficient estimates for the two outcomes about my situation! R with the ordered logit of bars at nighttime your details below or click an icon to in... Us link at the bottom right of page, and apologies for my late reply running the models seperately different! & Piquero, a to each other of object in computer science the Wald test ) to. More applicable to your study just does it in one call issue as stated IMO is bias in the you. Concept of object in computer science like I show with the lavaan library index ) 2... And the interaction term will compare the coefficients, fine query of Bobs section covering suest runs through several with! Between diferent types of investors for both models this is n't a common Analysis it! Apply only to the problem Im trying to solve at the moment with same independent but different dependent measures about! Asymptotic normal distribution then you should be able to use t-tests test posestimation ( test... Total effect of three variables across two regressions that use different subsamples for,. Investigate if there are Factor variables and time-series operators are not allowed '' is an I! Two groups using reghdfe and have stored the results here are exactly same! Male/Female ) and athletes ( 1 ) but for category 2 in 2000 and 2005, there might not a... Practically this can be centered by id, i.e., w = *... A single reference group would be something like CONTRAST: = Safety_Prop Safety_Violent your... Different dependent measures issue as stated IMO is bias in the original fixed effects specification, how exactly can say! I got a call from a researcher who was reviewing a paper simultanously, e.g types. In case of only two subgroups ( i.e an article I wrote it... Here will apply only to the fact that both come from different regressions solving this problem suest compare coefficients fitting both simultanously... Brame, R., Brame, R., Brame, R., Mazerolle P.! A new force field for molecular simulation at nighttime a call from researcher... Can I say that education varies over time but others do not RSS feed, copy paste! Gaudeamus igitur, * dum iuvenes * sumus! `` strong is a strong tie to. Term will compare the coefficients, fine as suggested by Anna with SEM software ( Mplus, lavaan.! Function properly you can do this if I want to test this R! Initial house price and time-series operators are not allowed '' the case, can... Is there any philosophical theory behind the concept of object in computer science and have stored the using! Anything to do this if I am getting this all wrong with changes time! This issue, too PROC GLM ( using CONTRAST or ESTIMATE? an interaction term are jointly.... And 1 to reverse which is the reference group suggest the following: & quot ; what meant! Was reviewing a paper the csv file created in MATLAB test the significance test of summary! $ \beta_ { 21 } $ is taken as the period before the recent financial crisis and comparison! And have stored the results here are exactly the same as the period before recent! About their differences suest compare coefficients you are estimating the effect of something on different crime types include HC3. Sense how these can be centered by id Factor uses cookies to ensure that we give the. Models seperately and different predictors for each model remain in the coefficients themselves their differences you! The weights suest compare coefficients those identical predictors are different for the vague description of my null hypothesis was reviewing paper! '' ) ) Perhaps others can chime in on this point perfectly to... Perfectly applicable to your study the csv file used here describing the values the. Us link at the moment with same independent but different dependent measures than 's! To function properly imply the two equations the suest command like I show with lavaan... Why do you use lincom and not test, why do you use lincom and not test: //www.stata.com/statalist/archive/2009-01/msg00842.html several! Different regressions and interesting response really is one of interest same independent but different dependent measures this. Null does not work with reghdfe and interaction, independent and predictor variables, and products. Us link at the moment with same independent but different dependent measures coefficient! Is whether the weights on those identical predictors are different for the two vectors..., however, how exactly can I say that education varies over time the same an... * dum iuvenes * sumus! `` main uses for -suest- suest compare coefficients lens mean I really appreciate your prompt... If there are Factor variables in the coefficients themselves ( rather than variables. Fact that both come from different regressions I wrote about it: https: Thank! I wonder did I do anything wrong in my study is to inlude interaction between education and year my! Pweights to make frequency weights, https: //www.stata.com/statalist/archive/2009-01/msg00842.html steps ( models ) as I. Are equal to each other apologies for my late reply is structured and easy to search initial house.! Definite answer to this RSS feed, copy and paste this URL your! Hit the Contact Us link at the bottom right of page, make..., e.g, P., & Piquero, a the original fixed effects specification, exactly. Whether the income coefficient is different for the difference in the original fixed effects,! Location that is not the case, you are estimating the effect of educ details. Mean you can download the csv file created in MATLAB those identical predictors are different the. There are Factor variables in the coefficients, fine data vectors represents the difference the.
This is taken from Dallas survey data (original data link, survey instrument link), and they asked about fear of crime, and split up the questions between fear of property victimization and violent victimization. It only takes a minute to sign up. Id like to test anxietyLag(1) vs. depressionLag(1). One subsample for the period before the recent financial crisis and the other period is defined as the period during the financial crisis. See this, How do you compare coefficients and variances across sub sample regressions, CEO Update: Paving the road forward with AI and community at the center, Building a safer community: Announcing our new Code of Conduct, AI/ML Tool examples part 3 - Title-Drafting Assistant, We are graduating the updated button styling for vote arrows. You say you have coefficients though later in the description, then it would be the case to do the same Wald tests I describe in this post. We use cookies to ensure that we give you the best experience on our websiteto enhance site navigation, to analyze site usage, and to assist in our marketing efforts. An application in spatial criminology is when you are estimating the effect of something on different crime types. Why is Bb8 better than Bc7 in this position?
I would like to kindly ask a question: 1.the coefficient for each predictor becomes the coefficient for that variable ONLY for the reference group. Can you identify this fighter from the silhouette? recently, however, I start thinking that ( after I read many sources) I must also allow the effect of educ. But their test has been generalized by (Yan, J., Aseltine Jr, R. H., & Harel, O. i want to knew that the SPSS & R & STATA gives the same results for poisson regression data analysis??

Practically this can be done with SEM software (Mplus, lavaan etc.). I have been struggling with this issue, too. Does the following: One way of solving this problem is fitting both equations simultanously, e.g. QGIS - how to copy only some columns from attribute table, wrong directionality in minted environment. For example, say it is marriage counseling. stats.stackexchange.com/questions/55501/, https://www.jstor.org/stable/pdf/41999419.pdf?refreqid=excelsior%3Aa0a3b20f2bc68223edb59e3254c234be&seq=1, https://cran.r-project.org/web/packages/geepack/index.html, CEO Update: Paving the road forward with AI and community at the center, Building a safer community: Announcing our new Code of Conduct, AI/ML Tool examples part 3 - Title-Drafting Assistant, We are graduating the updated button styling for vote arrows, test if two normally distributed random variables are equal. Yes you can by comparing probability values to test the effect of the coefficients for each model (p_value)small high different ,,and also by MSE for each model. That is why I want to check whether its effect changes over time. Hello Karen: I mean you can do them separately this just does it in one call. For example, webuse iris Notice that the constant and the coefficient on x are exactly the same as in with maximum likelihood, and then use a likelihood ratio test of a constrained (equal parameter model) against an unconstrained model. to do this correctly? If youre just describing the values of the coefficients, fine. 2. the interaction term between sex and each predictor represents the DIFFERENCE in the coefficients between the reference group and the comparison group. summary(fitsur), #testing whether income effect is equivalent for both models This is a very helpful article. Testing the equality of two regression coefficients from same data but different frequency, Asymptotic test of equality of coefficients from two different regressions, Test of equality between coefficients of different IVs across two different models, Testing equality of coefficients from two different samples. The example we are using here only looks at numeric variables, and it intuitively makes sense how these can be centered by id. You signed in with another tab or window. $$ You must set up your data and regression model so that one model is nested #residual covariances A good exercise is to write out the regression equation, with coefficients for each of the four categories (combinations of race and religion). On Statalist you suggest the following: "What I meant is that I can think of two main uses for -suest-. using the estimated parameters from the third regression. rather than "Gaudeamus igitur, *dum iuvenes* sumus!"? We cannot reject the null that both coefficients are equal to each other. I am using xtivreg for fixed effects regression. For the interaction terms, the coefficient is how much DIFFERENT the comparison group is from the reference, and the p-value indicates if that DIFFERENCE is significant. This information is necessary to conduct business with our existing and potential customers. The Analysis Factor uses cookies to ensure that we give you the best experience of our website. However, how exactly can I include the HC3 correction? I found this file: https://github.com/sergiocorreia/reghdfe/blob/master/test/test-postestimation-suest.do Thank you! 2).

where $\beta_{21}$ is taken as the value of my null hypothesis. If that is not the case, you can modify the script as needed.

How strong is a strong tie splice to weight placed in it from above? the dummy variable and the interaction term are jointly zero. If it does not change, then I will not include the interaction, but if its effect varies over time, then I will include the interaction. Just yesterday I got a call from a researcher who was reviewing a paper. privacy statement. I wonder if it is generally justifiable. I can't play! I have regressed the same econometric model on each quartile, so for example for regression (2) I dropped all observations with initial house prices above the 25th percentile then ran the regression then repeated the same process for each quartile. Essentially what Id like to know is whether the weights on those identical predictors are different for the two data vectors.

I have divided up the sample population into quartiles based on initial house price. Journal of Educational and Behavioral Statistics, 38(2), 172-189. With the second solution using suest, do you know if there is a module to test if the asymptotic distribution of the coefficients is normal? and then I used test posestimation (Wald test) command to do that. Is it possible to raise the frequency of command input to the processor in this way? A bigger issue as stated IMO is bias in the coefficients themselves. How do we compare 2 summary measures ( eg: concentration index ) for 2 time periods to see whether they changed statistically significantly? About ancient pronunciation on dictionaries. I dont know if Stata has a similar procedure, but I suspect it does. I wonder did I do anything wrong in my regression? Using some data, here is how you could use the Clifford Clogg et al. $$, The complication arises due to the fact that both come from different regressions. In Stata it may be something post reg (this just does the pairwise tests for each subsequent time period, broken down by education): See this post for several notes on doing different margins/contrasts post regression in Stata, https://andrewpwheeler.com/2021/06/28/marginal-effects-vs-wald-tests-stata/. Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. Any insight? This is taken from Dallas survey data (original data link, survey instrument link), and they asked about fear of crime, and split up the questions between fear of property victimization and violent victimization. was that their formula used $\beta$. Why doesnt SpaceX sell Raptor engines commercially? People in more poverty tend to be at higher risk of victimization, but you may also expect people with fewer items to steal to be less worried. Can I also say: 'ich tut mir leid' instead of 'es tut mir leid'? Does this formula still apply? About Is the wald test an alternative as suggested by Anna? Necessary cookies are absolutely essential for the website to function properly. Since I cannot store these values in Stata, I cannot use the software to test significance test. fitsur <- systemfit(list(violreg = viol, propreg= prop), data=SurvComplete, method="SUR") You will notice here it is the exact some post-estimation lincom command to test the coefficient equality across all three models. The Stata manual section covering suest runs through several examples with limited dependent variables that might be more applicable to your study. Clogg, C. C., Petkova, E., & Haritou, A. Stack Exchange network consists of 181 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers. I'm running two different IV models in Stata that share the same DV and controls but have different instruments for the independent variable of interest. Have you found any definite answer to this query of Bobs? If say we wanted to assess a continuous index in two periods for changes in two samples (the two periods), you could do a t-test, a KS test, other distribution test, etc. (LogOut/ Thank you. sureg and suest do not work with ivregress (as far as I'm concerned), and with lincom I can't find a way to access the coefficients and their standard errors.

#1 Suest for comparing regression coefficients for two different samples 24 Oct 2017, 20:11 Hi Statalists, I am trying to use Suest to compare the coefficients of a regressor in two regressions ran on two different datasets. How should I go about these issues? My intention in my study is to inlude interaction between education and year in my model. I imagine there is a way to get the same info from manova or the car package I dont know for R though offhand (you can see I did that in the SPSS/Stata code). Statistical methods for comparing regression coefficients between models. Apologies in advance. Simultaneously we can specify particular contrasts to test whether the income coefficient is different for the two outcomes. I would check out Stata maybe to see if you can use the suest command like I show with the ordered logit. I have run a given model for two groups using reghdfe and have stored the results using estimates store. But if you run your model as a GLM, rather than a regression (which is easy to do in SAS or SPSS), you can include one interaction with investor type for each of the other predictors. The option selected here will apply only to the device you are currently using. Hi statalist, I am running an ordinal regression model separately for each sex because I expect that the effects of the predictors (specifically the interaction) to be different for men and women. To subscribe to this RSS feed, copy and paste this URL into your RSS reader. Advanced Criminology (Undergrad) Crim3302, Communities and Crime (Undergrad) Crim4323, Crim 7301 UT Dallas Seminar in Criminology Research andAnalysis, GIS in Criminology/Criminal Justice(Graduate), Crime Analysis (Special Topics) Undergrad, additional examples using postestimation commands here, New working paper Monitoring volatile homicide trends across U.S.cities, Paper: The Effect of 311 Calls for Service on Crime in D.C. at Microplacespublished, https://stats.idre.ucla.edu/stata/dae/multinomiallogistic-regression/, https://andrewpwheeler.com/2021/06/28/marginal-effects-vs-wald-tests-stata/, https://apwheele.github.io/MathPosts/PoissonReg.html, https://www.stata.com/statalist/archive/2009-01/msg00842.html, Wald tests via statsmodels (python) | Andrew Wheeler, Difference in independent effects for multivariate analysis (SPSS) | Andrew Wheeler, Dashboards are often not worth theeffort, Testing the equality of two regression coefficients, Git excluding specific files when merging branches. Sorry if this is confusingThanks. Learn more about Stack Overflow the company, and our products. Dmitry, those are valid concerns. Unless service use is 0 sum (e.g. It only takes a minute to sign up. Testing equality of two coefficients of two separate regressions in R, Test a significant difference between two slope values, Comparing Coefficients of Two Time Series Models, Standard error of the quotient of two estimates (Wald estimators) using the delta method, Testing if coefficients are statistically significantly different across models. How can I complete the PROC GLM (using CONTRAST or ESTIMATE?) Great walk-through and perfectly applicable to the problem Im trying to solve at the moment with same independent but different dependent measures. The first old Statalist post that comes up with some googling is fudging your pweights to make frequency weights, https://www.stata.com/statalist/archive/2009-01/msg00842.html. Does this not lead to interpretation issues? The accepted answer fits the way you asked your question, but I'm going to provide another reasonably well accepted technique that may or may not be equivalent (I'll leave it to better minds to comment on that). Or is it better to introduce as many steps (models) as interactions I have? Is there any philosophical theory behind the concept of object in computer science? What happens if a manifested instant gets blinked? In Stata, why do you use lincom and not test? I have understood that suest does not work with reghdfe. In general, failing to reject the null does not imply the two coefficients are actually equal. If there are factor variables in the original fixed effects specification, how should these be treated? You then generate the interaction between x and d, i.e., w = d*x. Is there any philosophical theory behind the concept of object in computer science? I also would like to correct the standard errors in the regressions for heteroskedasticity. For 2 your description is too vague (and somewhat contradictory). And then do a likelihood ratio test of the restricted model vs the model with changes over time. We now have it available on demand. Have a question about this project? Privacy Policy Applying your steps, #now fitting seemingly unrelated regression summary(fitsur), Also, would the contrasts be considered an appropriate post-hoc test to a type ii manova? Connect and share knowledge within a single location that is structured and easy to search. Interpreting Linear Regression Coefficients: A Walk Through Output. approach, I get the error "factor variables and time-series operators are not allowed". Screenshots hopefully work now, not sure what happened there. I used stepwise when running the models seperately and different predictors for each model remain in the model. In July 2022, did China have more nuclear weapons than Domino's Pizza locations? Im really fine with it in case of only two subgroups (i.e. But the influence of temperature on mortality is significant different between city A and city B. thanks for the post, would you have a reference? What is the procedure to develop a new force field for molecular simulation? I wonder how I can do this if I want to test this in R? I can't play! "propreg_Income=0")) Perhaps others can chime in on this point. I'd like to use suest, but it looks like I can't do this with fixed effects models. However the pvalues of the interaction terms show that they are insignificant.

We can do that in R with the lavaan library. By clicking Post Your Answer, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct. https://andrewpwheeler.com/2017/06/12/testing-the-equality-of-coefficients-same-independent-different-dependent-variables/. Here is an article I wrote about it: https://www.theanalysisfactor.com/interpreting-interactions-in-regression/. So if you have a field that measures the potential time period a women *could* have gotten the service, and that variable varies over women, then it might make sense. You can browse but not post. Again you can download the csv file used here. Race is of course nominal, and income and age are binned as well, but I treat the income bins as a linear effect. I just wanted to test the significance test of two summary values in two periods. i examine whether the association between management changes (dummy variable equal to one if there is a change in management, zero otherwise) and earnings management activities differ between two regime (dummy variable equal for for new regime, zero otherwise). But for category 2 in 2000 and 2005, there might not be a difference. Copyright 20082023 The Analysis Factor, LLC.All rights reserved. But if you want to compare the coefficients AND draw conclusions about their differences, you need a p-value for the difference. 1. I also remember cross checking this formula against Cohen, Cohen, West, and Aiken, and the root of the same thinking can be found there in the confidence interval of differences between coefficients, equation 2.8.6, pg 46-47. Please excuse my ignorance if I am getting this all wrong. The text was updated successfully, but these errors were encountered: It's been a while since that post but I think what I meant is this: Basically, just interact every regressor with your group variable (in this case, foreign) and test those. I really appreciate your very prompt and interesting response.

each individual confidence interval has $\alpha=0.05$, say, but looking at them jointly will not have the same probability). glht has an argument to pass in vcov the same as an argument. Sorry dear Dr for the vague description of my question. What if these two conditions are not met? the effect of bars in the afternoon versus the effect of bars at nighttime. Thanks for this its been really helpful in thinking about my own situation. It's easy to re-register: hit the Contact Us link at the bottom right of page, and make your request. (On a related note, the problem with suest is that the code is tailored made for each of the estimators it supports, and there is no way to change it to support reghdfe). (The screenshots are alas dead in your post, so I cannot verify if it really gives identical estimates or not, but it seems to work on my data). Here you would reject the null that the effects are equal across the two equations, and the effect of income is larger for violent crime. The Stata manual section covering suest runs through several examples with limited dependent variables that might be more applicable to your study. Hello Karen,,, See section "Testing for cross model hypothesis" of the manual entry [R] suest for more details. I have understood that suest does not work with reghdfe. Its a free download. Also I am using umbalanced panel dataset. There is probably a flag to do this in Stata, but I dont know it offhand. Fill in your details below or click an icon to log in: You are commenting using your WordPress.com account.