Dear all, I have conducted a double-bounded contingent valuation study and analysed it in Stata with Lopez-Feldmann's (2012) doubleb command. The resulting coefficients are confusing me. I am not sure whether they are standardised or unstandardised and this is not helping with my conclusions. I know that probit regression usually returns standardised coefficients, so I would not be able to say that a 1 unit increase in the IV results in a (coef.) increase in the DV. However, since the Lopez-Feldmann model is a variation of probit regression and his explanation of it shows the coefficients to be unstandardised(see below), I have interpreted my results in this way. But now I am uncertain and need to defend my results on Wednesday. From Lopez-Feldmann (2012) - https://mpra.ub.uni-muenchen.de/4101...aper_41018.pdf The last output before the references shows the difference between male and female WTP to be exactly that of the coefficient : so this would mean the coefficient is standardised, right? I have now interpreted all my coefficients (some linear, some categorical some nominal) as unstandardised. Is this correct? If so, how can I argue that it is correct that they are unstandardised although they are usually standardised in probit? Here is my Stata output: initial: log likelihood = -<inf> (could not be evaluated) feasible: log likelihood = -6032.3338 rescale: log likelihood = -466.69701 rescale eq: log likelihood = -464.61633 Iteration 0: log likelihood = -464.61633 (not concave) Iteration 1: log likelihood = -447.67555 (not concave) Iteration 2: log likelihood = -410.05117 (not concave) Iteration 3: log likelihood = -402.94301 (not concave) Iteration 4: log likelihood = -399.09683 Iteration 5: log likelihood = -375.40299 Iteration 6: log likelihood = -364.54334 Iteration 7: log likelihood = -361.66365 Iteration 8: log likelihood = -361.64446 Iteration 9: log likelihood = -361.64445 Number of obs = 137 Wald chi2(8) = 53.00 Log likelihood = -361.64445 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- Beta | age | -.3450866 .1649973 -2.09 0.036 -.6684753 -.0216978 gender | 7.255382 3.653988 1.99 0.047 .0936969 14.41707 income | -.0000227 .0020633 -0.01 0.991 -.0040666 .0040212 org | 2.666961 1.273459 2.09 0.036 .1710266 5.162896 qualU | 19.41732 3.6545 5.31 0.000 12.25463 26.58001 self | -.0649317 .1229359 -0.53 0.597 -.3058816 .1760183 env | .2843504 .1289765 2.20 0.027 .0315611 .5371396 soc | .1056503 .1646891 0.64 0.521 -.2171343 .4284349 _cons | 44.36785 11.70759 3.79 0.000 21.4214 67.3143 -------------+---------------------------------------------------------------- Sigma | _cons | 16.75996 1.485025 11.29 0.000 13.84937 19.67056 ------------------------------------------------------------------------------ First-Bid Variable: bid1 Second-Bid Variable: bid2 First-Response Dummy Variable: answer1 Second-Response Dummy Variable: answer2 So as an example: I said that an increase from one category of organic buying frequency to the next (org) causes a 2.67 percentage point increase in Willingness to Pay. Any help would be very much appreciated Thanks Joe
Related Posts with Probit regression coefficients
Using estimates from MI regression in Forecast functionI would like to use estimates from a regression (that used 'multiple imputation') in the 'forecast' …
Frequency table combining variable values id spec2010 grade2010 spec2011 grade2011 spec2012 grade2012 spec2013 grade2013 1 ren ct1 ren c…
How to generate the number of schools a student is eligible for within his own state and year? I have the following dataset: school id score state year university_a 1 150 Cal 2000 university_a 2…
Check if variable at datetime equal variable at datetime + 24 hoursGood day, I have two variables (among others): One variable is a binary variable 0 or 1 (sleepwake)…
destring date variableHi, I have trouble converting a string date (dd.mm.yy.) to a stata date variable. I have tried this…
Subscribe to:
Post Comments (Atom)
0 Response to Probit regression coefficients
Post a Comment