Search Results

Author like: Teachman, Jay D
Resulting in 1 citation.
1. Akerlof, George A.
Main, Brian G.
Maximum Likelihood Estimation with Pooled Observations: An Example from Labor Economics
International Economic Review 21,3 (October 1980): 507-515.
Also: http://www.jstor.org/stable/2526348
Cohort(s): Older Men
Publisher: U.S. International Trade Commission
Keyword(s): Data Quality/Consistency; Employment; Longitudinal Data Sets; Research Methodology; Unemployment

The difficulty of pooled observations in a data base arises when, for each sample point, some characteristic is measured exactly up to a certain level, but no record is made if the characteristic should exceed that boundary. This poses obvious problems if it is desired to perform a maximum likelihood estimate of the probability of an event. It is shown that whereas most researchers use ad hoc methods, it is possible to solve this problem by a non-ad hoc procedure which is illustrated by an example from labor economics. The example arises from a study of the weekly probability of transition from employment to unemployment using the NLS of Older Men. It is assumed that the probability of predicting whether an individual will be employed or unemployed in the following week can be represented as a logistic function of personal characteristics. A method is illustrated by which pooled data can be incorporated into maximum likelihood estimation in an exact fashion.
Bibliography Citation
Akerlof, George A. and Brian G. Main. "Maximum Likelihood Estimation with Pooled Observations: An Example from Labor Economics." International Economic Review 21,3 (October 1980): 507-515.