Modeling change: A gentle introduction to cross-lagged and latent growth curve approach
Time & Location
26 Sept, 09:00 – 27 Sept, 19:00
Zagreb, Zagreb, Croatia
About the Event
We would like to invite you to join us for the second free, two-day statistical course offered as part of the COORDINATE project. The goal of COORDINATE is to mobilise the community of researchers and organisations to advance the coordinated development of the birth cohort and survey research on children’s wellbeing in Europe.
The statistical course aims to provide practical knowledge on analysing longitudinal data, focusing on: longitudinal structural equation modelling, cross-lagged mediation analysis, unconditional and conditional latent growth curve analysis. Basic understanding of regression and factor analysis is required. A dataset will be provided for exercises, but participants can also use their own data. JASP (with lavaan syntax) and IBM SPSS AMOS statistical software packages will be used in the course.
The course is open to academic researchers from PhD students to full professors, policy practitioners, and other researchers or analysts working in the EU countries and associated states.
Day 1: Cross-lagged mediation analysis of longitudinal (panel) data
- Recapitulation of the key conceptual and methodological issues in longitudinal data analysis
- Structural equation modeling primer
- Cross-lagged analysis of 2- and 3-wave data
- Mediation analysis with cross-lagged models
Day 2: Latent growth curve (LGC) modeling
- Concept, use, and advantages of LGC
- Steps in LGC modeling
- Unconditional LGC; curve specification, estimation, and interpretation
- From one-domain to multi-domain conditional LGC models
At the end of this brief course, participants will:
- be able to test measurement invariance on longitudinal data
- be able to check for attrition bias and consider the effect of missing information
- be able to apply cross-lagged models on longitudinal data to estimate causal relationships
- understand the basics of mediation analyses and testing of indirect effects
- be able carry out mediation analysis with cross-lagged models using 2-wave (half-longitudinal mediation) and 3-wave data
- be able to carry out a LGC model with specified and unspecified change
- be able to carry out and correctly interpret a conditional LGC model
- understand in which cases cross-lagged and LGC modeling would be preferable to longitudinal (multilevel) regression modeling.
Preliminary course schedule
Day 1: September 26, 2023
09:00 – 09:40 Introduction; Lecture 1: Key issues in longitudinal data analyses
09:40 – 10:15 Lecture 2: Structural equation modeling primer
10:30 – 13:00 Lecture 3: Cross-lagged models and mediation analysis
13:00 – 14:00 Lunch
14:00 – 17:00 Practical exercises on cross-lagged models and mediation analysis in JASP and lavaan (with a break)
Day 2: September 27, 2023
09:00 – 10:30 Lecture 4: LGC modeling
10:45 – 13:00 Practical exercises in LGC modeling I
13:00 – 14:00 Lunch
14:00 – 17:00 Practical exercises in LGC modelingII
About the lecturers
Mitja Ružojčić, PhD is post-doctoral researcher at the Department of Psychology at the Faculty of Humanities and Social Sciences of the University of Zagreb.
Aleksandar Štulhofer, PhD is full professor of Sociology at the Faculty of Humanities and Social Sciences of the University of Zagreb.
To be eligible for participation in this course, applicants have to meet the following criteria:
- The applicant’s interests are in the fields-of-interest of the COORDINATE project e.g., child and youth well-being.
- The applicant must have a contract or affiliation with a recognized academic or public institution, or a not-for-profit organisation registered in an EU member state or in an associated country.
- Applicants not working in an EU or associated country can apply and attend the course but are not eligible for financial support for travel and accommodation.
- Researchers from COORDINATE partner institutions who are not involved in the COORDINATE project are eligible to apply. However, as the objective of the project is to widen the network of researchers using child well-being data, only a limited number of places will be awarded to applicants from COORDINATE partner institutions.
- PhD students are eligible to apply. However, masters or bachelors students are not eligible.
To apply, click here. The deadline for the application submission is 19th June 2023.
Priority will be given to established researchers in the field, and to applicants from low GDP countries. Gender balance and geographical diversity will also be taken into account.
The applications will be reviewed by a committee, which will score each application based on the information provided in the application form.
Applicants will be notified about the decision by 30th June 2023.
A limited number of bursaries (up to €500) to support travel and accommodation will be available to selected participants. Additionally, two lunches and one dinner will be provided for all attendees.
If you have any questions, please contact firstname.lastname@example.org
The course is organised by the Croatian Social Science Data Archive, University of Zagreb Faculty of Humanities and Social Sciences, Croatia, within the COhortcOmmunity Research and Development Infrastructure Network for Access Throughout Europe project.