UDIFF: Stata module to estimate the generalized unidiff model for individual-level data

Jann, Ben; Seiler, Simon (2019). UDIFF: Stata module to estimate the generalized unidiff model for individual-level data. [Software & Other Digital Items]

[img] Archive
udiff.zip - Published Version
Available under License BORIS Standard License.

Download (19kB)

udiff estimates parameters of the so-called unidiff model (Erikson and Goldthorpe 1992: The Constant Flux. Oxford University Press), also known as the log-multiplicative layer effect model (Xie 1992: The Log-Multiplicative Layer Effect Model for Comparing Mobility Tables. American Sociological Review 57: 380–395), which is often used to study differences in intergenerational class mobility between birth cohorts or countries. The original unidiff model has been expressed as a log-linear model of cell frequencies in a three-way contingency table (origin by destination by cohort or country). The model, however, can also be expressed at the individual-level (similar to a multinomial logit model). udiff estimates such a re-expressed unidiff model for individual-level data. Furthermore, it generalizes the model to allow for multiple layers and non-categorical predictors. For an implementation of the classic log-linear unidiff model for aggregate data see Pisati (2000: sg142: Uniform layer effect models for the analysis of differences in two-way associations. Stata Technical Bulletin 55: 33-47).

Item Type:

Software & Other Digital Items


03 Faculty of Business, Economics and Social Sciences > Social Sciences > Institute of Sociology

UniBE Contributor:

Jann, Ben and Seiler, Simon


300 Social sciences, sociology & anthropology


Statistical Software Components


Boston College Department of Economics




Ben Jann

Date Deposited:

19 Nov 2019 11:47

Last Modified:

20 Jan 2020 13:06

Related URLs:

Additional Information:

Statistical Software Components S458711





Actions (login required)

Edit item Edit item
Provide Feedback