Propensity Score Matching in Stata - psmatch2

[@marianabenitezrubiano9207]

Thanks a lot, amazing professor. Greetings from Colombia.

[@asmazafar6299]

Thanks, Chris! For this super helpful tutorial!

[@osmangani]

Ever best video on PSM. We want more from you.

[@getachewgenetu4916]

thank you for very interesting and educative video. from Ethiopia.

[@berthakadawayula9510]

I have learnt alot. Please make a video on the t effects

[@arzanamyderrizi1970]

Thank you, your video is super useful!

[@MUBASIRUASAFELamidi]

Hi, thank you for this educative video. Please how can I get the dataset you use in this presentation to be able to replicate your examples. Thank you for your help.

[@parulgupta788]

So the headstart method actually reduced math achievement??

[@RNHD-se4dk]

Is there a way to extract the dataset specifically from the untreated (but matched) population??

[@felipemunozcastillo2387]

Thanks, Chris!

[@giorgioricciardi7048]

Amazing video. I have a question. When I perform PSM, below the regression table appears a line that says: " Note: 2 failures and 0 successes completely determined". Do you know what does it mean?. Thank you.

[@sahilnarang2986]

This is called a problem of quasi or complete separation. You must be having binary variable as dependent variable, which is completely predicted by one of the regressors. Understand in this way: I have a regressor in which dependent variable taken constant value say either only zero or only 1. This means that regressor is able to determine the success rate (if dv=1) and failure rate if dv=0 corresponding to that regressor.

[@NASIRABBASKHAN]

wonderful video! thanks for uploading, can you please make a video on calculating marginal effects and partial elasticity in a binary logistics reg. model in stata

[@sabaaslam4766]

Hi, it is showing ''probit failed to estimate propensity scores; computations cannot proceed''. Do you have any idea about this error?

[@renemarcelofernandezmerlo8278]

Hi and nice video, i have a question, how do you interpret T-stat from ATT and if it's not significant what it means?

[@sahilnarang2986]

If not significant that means no significant difference between matched control and treatment firms with respect to the outcome variable.

[@sampurnakundu2612]

I tried to use the psgraph command after the psmatch2 estimates, but it coming like 'no data is mi set', can't understand what it means

[@christina1302]

Hi, thank you for the video I have found it really helpful with my project! Can I just check - above, in order to evaluate the outcome we are simply taking the mean difference between treated and untreated units (assuming that we have found balance, of course)? Thanks!

[@tmeyers6873]

Hi. I was wondering if you could provide information on how to interpret the results from the psmatch2 procedure for significance? How does one interpret the t-tests for the matched ATT?

[@rahulkamal8431]

hey, even I have the same question. pls let me know if u have the answer

[@melissa3232]

Hello. I was wondering if you found a response to this question. I am also wondering how to report the results with a significant att?

[@sahilnarang2986]

@@melissa3232 This should be very simple. No significance (via t test) means no significant difference between matched control firms and the treatment group. Difference or no difference, we are comparing with respect to the outcome variable.

[@roypeijen]

Dear. F. Chris Curran, it might be a stupid question, though, I would like to ask. It isn't a problem that some covariates are dropped in the matching procedure? It just indicates that the ' treatment group' does not score on those indicators (for instance, value=0) ? Am I right?

[@sckang4692]

hi, I followed your guide, however, I cannot find which variables are chosen control variables through PMS. For example, I did Nearest-neighbor matching and the result shows several control variables were chosen. But which variables were chosen among all control variables? There is no result showing the chosen variables. I want to regression after PMS. however, your guide seems not to belong to how I can find the PMS result from variables data related to which one is chosen.

[@hassanzaidi7081]

Hi, I have a question. How can we apply if condition in scoring and matching? For example, if we have a dataset of 5 years but we want to limit the treated variable for just two years.

[@yohannesaycheluhim835]

*evaluate match graphically . psgraph unknown egen function sum() r(133); how ca I fix thi problem?

[@saminaaftab3088]

Evaluate match graphically is showing error kindly help out

[@anjalipant5350]

Hi Curran! Excellent presentation. I wish to replicate this study for practice. From where can I get this dataset?

[@md.arrahman7125]

Dr. Thanks a lot, would you please help me that how calculate PSM-DID for long panel data?

[@md.arrahman7125]

@C. Comploj not yet Dr.

[@drzeeshanfareed2404]

Excellent work within a seconds all done. Sir My DV is "Risk", IV is "CG" and controls are "C1" "C2" "C3".. I have Binary variable which dummy. 1 represent public firms and 0 Private firms..... Please how to calculate PSM with nearest matching. Please help me. Thanks

[@hailyetekle500]

please can you help me not work for me ??? please send me dataset

[@mathewwade9099]

Thanks for the amazing video. Does anyone know how I could use psgraph in RStudio?

[@fajriahjulianti2484]

Hi! Do we have to use different command if using psm for panel data? Thank u

[@tongtonghong5653]

Hi! I also have the same problem as you do. I also want to do some psm using panel data. Did you have your problem solved?

[@suhailpc4282]

hello madam, i saw your query about panel data set...i hope you solved your problem. now iam facing the same problem, could you help me to solve this problem....could you pls give your conatcts or mail ...my e-mail id... [e-mail]...watsup: 9656039836