News Statistics1

Published on June 17th, 2013

0

Advanced Social Statistics: Ph.D.-Course at NTNU

The department of Sociology and Political Science at the Norwegian University of Science and Technology (NTNU) announces a Ph.D.-course on advanced statistical methods. The course is also open for Master students, and takes place on August 12–16.

The overall theme of this course is the study of longitudinal data in social science and medicine. This includes survival modeling, how to treat time-series cross-section data, and multilevel modeling.

 

Survival Analysis

A researcher would choose to employ this type of modeling when he or she is not only interested in whether or not an event happens, but also when it happens. It presupposes longitudinal data, that is, a study of correlations which involves repeated observations over long periods of time.

Many social-science phenomena are about causal relationships, that is, that they have a progress over time. Such causal connections can best be studied using experiments, however this approach have limited usefulness in many cases. The best non-experimental technique to study these processes of causality is by the use of survival analysis.

This type of modeling is also used by researchers within the field of medicine. The data may be structured in the way shown in the table presented below.

survival_time

event

women

treatment

6.00

1.00

.0

1.00

19.00

.0

1.00

.0

32.00

1.00

.0

1.00

42.00

1.00

.0

.0

42.00

1.00

.0

1.00

 

Time-series cross-section methods

When employing these methods, the data consist of comparable time-series (TSCS) data observed on a variety of units, as shown in the table shown below.

Country-years
Sweden 1980
Sweden 1981
Sweden 1982
Sweden 1983
Sweden 1984
Sweden 1985
Norway 1980
Norway 1981
Norway 1982
Norway 1983
Norway 1984
Norway 1985

This resembles panel data, where a large number of units are observed over a small number of waves (or interviews). However, the structure is quite the opposite; with TSCS-data we have relatively few units and long time-series.

TSCS data are common in the study of political economy as well as in conflict studies.

 

Multilevel modeling

These models involve data that are ordered hierarchically, that is, some units of analysis are considered a subset of other units. It could for example be that people are a subset of countries. The object of a multilevel analysis is to account for variance in a dependent variable measured at the lowest level, by investigating information from all levels of analysis.

 

For further information on the course, see:

http://www.ntnu.no/documents/10455/21226680/SOS8515+Avansert+statistisk+dataanalyse+i+samfunnsvitenskap+H13.pdf

 

Cover photo by U.S. Geological Survey

 

Tags: , ,



Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Back to Top ↑