Oral Presentation Paul Campbell - Tracing time: Using longitudinal data to understand changes in individual characteristics and circumstances, and make inferences about population-level change


Population-level change is driven by individual-level decisions. The advent of a range of longitudinal datasets, such as panel data and linked data, has made it possible to study change at the individual level in much richer detail. In this presentation, I will first provide a broad overview of the five studies that comprise my PhD. I will introduce four themes that connect these studies: personality and individual differences, demographic processes, longitudinal analysis, and novel data sources. I will describe two related studies in detail. Both studies use the Household Income and Labour Dynamics in Australia (HILDA) survey to study personality and life events through time. The HILDA data contains measures of the Big 5 Personality Traits, the dominant theory in trait psychology, and I will introduce these measures before discussing my analysis.

The first study assesses the relationship between personality and internal migration in Australia. The study considers migration events in a 12-month period, but further, migration intentions and certainty prior to this period, and the alignment between intentions and outcomes. In assessing this, the study focuses on the extent to which individual characteristics, particularly personality, explain migration outcomes. Results show small but significant effects of Extraversion and Openness on both migration and intention to migrate. Openness negatively predicted certainty around intention, and Conscientiousness related to the extent to which migration intentions align with outcomes for those who did migrate, while Extraversion predicted migration events among those without a prior intention to migrate.

The second study complements the first, focusing on the personality traits themselves: the extent to which they change over time, and the factors that explain such change. HILDA has measured personality four times, at 4-year intervals, giving a 4- 8- and 12-year interval to assess. Results suggest small but consistent changes over time. Age effects are more prevalent over the 12-year period than found in the earlier 4-year period, and some change is explained by life events, notably change in subjective health and life satisfaction. Change in personality traits, particularly change driven by life events typically predicted by scores on the Big 5, introduces endogeneity between personality and life events, and this should be considered when personality is used to predict life outcomes. Longitudinal analysis can also be used to help validate the reliability of cross-sectional measures, in this case the personality trait measures.

Taken together, these studies demonstrate that individual characteristics both effect and are affected by life circumstances. We can develop a more detailed insight into demographic processes by considering individual level constructs.

NOTE limited number of places for people who want to attend in person at the RSSS, Room 4.69, Seminar Room (New RSSS Building number 146). Allocation of physical attendance on first in basis, Email Naomi Snowball Naomi.Snowball@anu.edu.au>. People are welcome to join via Zoom. Details of Zoom available on request from Naomi.


Date & time

Mon 09 Nov 2020, 10–11am


RSSS Building 146 Ellery Crescent, Room 4.69, Seminar Room


Paul Campbell


Naomi Snowball
6125 1301


Updated:  26 October 2020/Responsible Officer:  Centre Director/Page Contact:  CASS Marketing & Communications