001) CCSR Doctoral Training College Courses

North West Doctoral Training Centre courses.

PLEASE NOTE: COURSES NEED TO BE PREBOOKED  AT THE CCSR WEBSITE BEFORE COMPLETING THIS PAYMENT.

36) Longitudinal Data Analysis

Course Summary

The importance of longitudinal analysis is becoming increasingly recognized across the social and medical sciences. However there are few analysts with the methodological skills to make appropriate use of longitudinal data. This course is intended to meet this need.

Teaching Methods

The course will comprise 3 consecutive days of teaching. The 3 days of  intensive training will be made up of lectures and computer-lab examples and exercises implemented with appropriate statistical software.

Course Aim

To provide students with the skills needed to design longitudinal research and conduct  appropriate analyses using longitudinal data including the use of random effects models for repeated measures data and event history analysis.

Objectives

• To gain facility in the concepts, designs and terms of longitudinal research;
• To be able to apply a range of different methods of longitudinal data analysis;
• To have a general understanding of how each method is representing longitudinal
processes;
• To be able to choose a design, appropriate method of analysis and plausible model for a
range of research questions.

Preliminary Reading

Dobson, A. (2002). An introduction to generalized linear models. Chapman and Hall.

Goldstein, H. (1995). Multilevel statistical models. London: Edward Arnold.

Hosmer DW, Lemeshow S (1999). Applied survival analysis. Wiley.

Singer, J,D and Willet, J,B (2003). Applied longitudinal data analysis. Modeling change and event occurrence, Oxford University Press.

Snijders, T.A.B. and Bosker, R.J. (1999). Multilevel analysis. London: Sage.

37) Causal Modelling in STATA

Course Requirements: Requires some previous experience with doing regression or statistical tests.

Familiarity with STATA is desirable (but not required).

Experience managing complex survey data is desirable (not required).

Course Summary

Many people move from simple regression to more complex causal modelling as their professional life develops.  This course introduces basic techniques that are helpful for making statistical inferences in the intermediate level models:  using a ‘long’ format panel data set; applying regression to panel data; drawing out causal interpretations. The course will also cover some causal concepts, describe statistical approaches to causal inference, give worked examples of regression models, and give hands-on experience in applied causal analysis using STATA.  The fixed-effect model that predicts the change in a dependent variable (over time) will be presented.  We contrast that model with pooled data and random-effects models.  Panel data, such as time-series, repeat measurements, and longitudinal studies, are provided to be used during the practical sessions.  No prior experience with such data is required.

Course Aims

At the end of this course, participants should have:

- An understanding of what fixed-effects regression models offer that is not found in standard regression.

- A knowledge of causal concepts such as endogeneity, confounding, intermediate variables, moderators and proximate cause.

- Hands-on experience using STATA for regression (this may be your frist STATA experience).

Course Programme

- Causal analysis with panel data.

- Panel regression (fixed effects and random effects models - practical session using xtrag command; post-estimation commands for xtreg variants).

- Opening the avenue to structural equation modelling using 2 or 3 equations (a practical session looking at the correlation and covariance matrices; discussion of how they relate to the output; a brief discussion of xtivreg and some alternatives to it).

- Intermediate, intervening, moderating and endogenous variables (a brief paper exercise in causal modelling with proximate and distal causes).

- Time, endogeneity, and the connections with event-history  analysis - a preview.

Overall, we discuss bias and mediating variables as they affect the regression slopes. A reference list s provided for further reading.

Preliminary Reading

- Fox, J. (2008) Applied regression analysis and generalised linear models, Los Angeles; London, Sage

- Mukherjee, C. H. White, et al. (1998) Econometrics and data analysis for developing countries, London, New York, Routledge

- StataCorp, L. P. (2007) Stata longitudinal/panel data: reference manual, College Station, Tex., Stata Press

- Wooldridge, J. M. (2003) Introductory econometrics: a modern approach, Australia; United Kingdom, South-Western College Publishing

- Gash, V. (2008) "Sacrificing their Careers for their Families? An Analysis of the Penalty to Motherhood in Europe" Retrieved June, 2009, http://www.ccsr.ac.uk/publications/working/2008-18.pdf