Application of Latent Markov Model in Social Work Research

In longitudinal modeling, we are often interested in the changes of a latent trait of a group of subjects over time. Such trait is often measured using a measurement model, which examines how items are related to the trait. Items are repeatedly administered to the same group over a certain period of time. For example, we may examine the changes of people’s welfare attitudes over 10 years by administering the same instrument to the same group one time a year for 10 years.

Contingent on research questions, data characteristics, and the assumed distribution of the response variables, researchers may select different models to analyze longitudinal data, such as the latent growth model and the mixture Markov model. Among these approaches, the latent Markov model (LMM) is particularly attractive. The LMM assumes the existence of a latent process which affects the distribution of the response variables. The latent process is further assumed to follow a Markov chain with a finite number of states, i.e., latent states. Thus, the LMM includes two components: (a) the measurement model which explains the conditional distribution of response variables given the latent process, and (b) the latent model which explains the distribution of the latent states.

The LMM is especially relevant to social work research, since it can easily be applied in longitudinal studies in which the construct under investigation is latent. Given the lack of studies based on the LMM in social work, this workshop will discuss the steps in building an LMM, strategies for model appraisal, and reporting guidelines based on real data.

Objectives Upon completion of the workshop, participants will be able to: 1. recognize how the latent Markov model is promising in modeling longitudinal data, 2. demonstrate understanding of the steps of model building for latent Markov models, and 3. use the LMest package in R to fit latent Markov models.

Content Based on the pedagogical techniques of learning-science-by-doing-science and problem-based learning, this workshop aims to address the following content: 1. summary of the characteristics and advantages of latent Markov models, 2. a concise and systematic set of procedures to establish latent Markov models, 3. application of the latent Markov models in social work research using a problem example based on real data, 4. a step-by-step demonstration of model construction, estimation, prediction, visualization, and diagnostics using the LMest package, and 5. how to draw inferences and conclusions.

Implications Overall, this focus on the latent Markov model can be invaluable to social work researchers, faculty, students, and professionals who wish to leverage state-of-the-art longitudinal models to derive more rigorous inferences and conclusions. Featuring real data analyzed with a cutting-edge statistical package, this workshop simplifies esoteric longitudinal models making them more applicable to social work research. By becoming familiar with the best practices in applying the latent Markov model, attendees can gain the expertise necessary to use a powerful statistical tool that can help enhance the examination of longitudinal changes in latent traits and advance research in social work.