0

I am trying to estimate the parameters of a theoretical model using a NLS (non linear least square) regression in Stata.The parameters are: alpha (sensitivity to the assets' returns of participants in our experiment), beta (sensitivity to risk) and gamma (sensitivity to the environment).

The two problems I encountered:

(1) The alpha parameter of my model cannot be estimated by the NLS regression (see screenshot #1 below).

(2) In the .do file, all variables are automatically marked blue except the beta variable which remains grayed out. I don't know the reason for this and I don't know if this is responsible for the problem (1). See screenshot #2 below.

Could you help me to estimate the three parameters of this model? Indeed, I have never worked with NLS regressions.

Screenshot #1

Screenshot #2

Nick Cox
  • 35,529
  • 6
  • 31
  • 47
Rassul Kalfane
  • 9
  • 1
  • We need a reproducible example. What is the point of declaring as many clusters as there are data points? – Nick Cox Nov 06 '21 at 13:49
  • For (2), `alpha` and `gamma` are not highlighted because they are "variables". They are highlighted because they are recognized commands/keywords in Stata (though `gamma` has long been deprecated I believe) whereas `beta` is not. – JR96 Nov 07 '21 at 22:52
  • @NickCox I declared clusters on the participants' code because they participate in 3 consecutive rounds. We have 3 data for each participant. – Rassul Kalfane Nov 14 '21 at 09:34
  • @JR96 I tried with other keywords (omega, lamda) and these words are still not highlighted in blue. Do you have any other ideas? Thank you. – Rassul Kalfane Nov 14 '21 at 09:34
  • But your first screenshot is showing as many clusters as observations used. – Nick Cox Nov 14 '21 at 12:46
  • Omega and lamda are not keywords. The moral of the story is .do files do not have the capability to recognize variable names in memory; the highlighting you see is coincidence. – JR96 Nov 14 '21 at 14:28

0 Answers0