Can Probit or Logit Models Be Used on Panel Data with 50 Subjects?

  • Context: Graduate 
  • Thread starter Thread starter monsmatglad
  • Start date Start date
  • Tags Tags
    Data Probability
Click For Summary

Discussion Overview

The discussion centers on the applicability of probit and logit models to panel data, specifically with a dataset comprising 50 subjects, each with 20 observations. Participants explore concerns related to sample size, bias, and the implications of using data from large public companies.

Discussion Character

  • Debate/contested

Main Points Raised

  • Some participants propose that probit or logit models can be used on panel data, but caution that the number of subjects may limit the statistical power.
  • Others argue that the selection criteria for the panel may lead to a biased sample, which could affect the validity of the results.
  • A participant mentions having 50 individuals and 20 observations each, seeking feedback on potential concerns regarding the use of this data for logit or probit modeling.
  • Another participant acknowledges that while 50 subjects might be sufficient for reasonable statistics, the bias towards large companies could restrict the generalizability of the findings to smaller companies.

Areas of Agreement / Disagreement

Participants express differing views on the adequacy of the sample size and the implications of bias in the dataset. There is no consensus on the overall suitability of the models for the given data.

Contextual Notes

The discussion highlights potential limitations related to sample bias and the representativeness of the data, but does not resolve these issues.

monsmatglad
Messages
75
Reaction score
0
Is it possible to use probit or logit models on panel data?

Mons
 
Physics news on Phys.org
monsmatglad said:
Is it possible to use probit or logit models on panel data?

Mons
Yes, but depending on the number of people there may not be enough data to do any meaningful statistics. Also, depending on the selection criteria the panel may constitute a very biased sample
 
I have about 50 different "individuals" and (optimally) 20 observations for each. the "individuals" are chosen based on a list of the 50 largest US, public companies at the time of the first observation. Do you see any immediate reasons for concern, based on this information (if I am using this data material for a logit or probit model)?

Mons
 
50 subjects should be enough for reasonable statistics. Your sample is highly biased towards large companies, so your conclusions should not be generalized to smaller companies.
 
  • Like
Likes   Reactions: ZeGato

Similar threads

Graduate "Population-averaged"regression on panel data using Stata
  • · Replies 1 ·
Replies
1
Views
2K
Undergrad What Are the Key Differences Between OLS and GLM Models in Statistical Analysis?
  • · Replies 3 ·
Replies
3
Views
3K
Undergrad What are the commonly used estimators in regression models?
  • · Replies 23 ·
Replies
23
Views
4K
Graduate Help with generalized linear model
  • · Replies 1 ·
Replies
1
Views
2K
Undergrad Time Series Models vs Models Used for Cross-sectional Data
  • · Replies 14 ·
Replies
14
Views
3K
Graduate Should I Ignore Data Driven Models or Use Bayesian Methods for Model Selection?
  • · Replies 17 ·
Replies
17
Views
2K
Undergrad Data Analaysis -- How to choose the best statistical model to use?
  • · Replies 5 ·
Replies
5
Views
2K
Undergrad Can these equations be solved analytically?
  • · Replies 6 ·
Replies
6
Views
2K
High School How can we accurately model data using curve fitting techniques?
  • · Replies 4 ·
Replies
4
Views
2K
High School It works but why? (Matching experimental data to a random equation)
  • · Replies 3 ·
Replies
3
Views
1K