13344 Chap.5 Longitudinal Designs and Change Over Time
Longitudinal Designs and Change Over Time
Measuring the same people again
The chapter defines longitudinal research, its uses for charting change and identifying causes, and the three conditions a causal claim needs: covariation, non-spuriousness and temporality.
Designs and models
It compares panel and multiple-cohort designs, explains period effects, attrition and the spacing of waves, and shows how to read cross-lagged panels and growth curves, the two analysis families the teaching names.
Using it in the essay
A longitudinal design is the natural second design in a title about associations, because it adds temporal order.
The chapter gives a worked reading of a two-wave panel and a short list of the design's advantages and disadvantages for an essay comparing designs.
What this chapter covers
- 01
Definition and purposes
- 02
Covariation, non-spuriousness and temporality
- 03
Panel and multiple-cohort designs
- 04
Period effects and attrition
- 05
Cross-lagged panel models
- 06
Growth curve models
Worked example · free
Which panel path tests whether depression predicts later insomnia
- 1The question concerns depression first and insomnia later, so the relevant path is d, from wave 1 depression to wave 2 insomnia.
- 1Its value, 0.05, is small, so earlier depression adds little to later insomnia once earlier insomnia is allowed for.
- 1Comparing it with path c, 0.25, suggests the association runs mainly from insomnia to depression.
Key terms
- Panel Study
- A Panel Study follows one group of participants over time with repeated measurement.
- Period Effect
- A Period Effect is a change caused by a historical event that affects everyone at once, which can mimic development.
- Sample Attrition
- Sample Attrition is the loss of participants over the waves of a study.
- Growth Model
- A Growth Model estimates each person's starting level and rate of change across three or more waves.
Longitudinal Designs and Change Over Time FAQ
Why can't comparing age groups replace a longitudinal study?
A comparison of age groups says nothing about how any individual changes, confuses age effects with cohort effects and gives no evidence about the order of cause and effect. Following the same people over time addresses all three.
How many waves does a growth model need?
The teaching states that growth models need at least three waves, with identical measures used at every wave. With four waves, the intercept loadings are 1, 1, 1, 1 and the slope loadings 0, 1, 2, 3.
How can a study reduce attrition?
The handout recommends collecting wide-ranging contact information at the start, building positive relationships and keeping regular contact, offering inducements for participation, and not over-burdening participants. It also advises checking who drops out.
What is a period effect?
A change caused by a historical event that affects everyone at once, such as an economic crash. In a single panel it can look like developmental change, which is why the strongest evidence for development is a pattern repeated across several cohorts recruited at different times.
What happens if waves are too far apart or too close together?
Waves that are too close may not allow enough time for meaningful change to appear. Waves that are too far apart make the study more sensitive to measurement error and may miss transitions in between, such as the move from primary to secondary school.
What is the difference between continuity and stability?
Continuity is the degree of consistency in group means over time, such as whether average low mood rises across adolescence. Stability is the consistency of individual differences, whether people keep their rank order. A measure can change on average while individuals stay in the same order.
Assessment move
Draw the four-box cross-lagged diagram from memory and label each path, then practise explaining a result in two sentences: what is stable and which crossing path is stronger. For each longitudinal study you read, note how many waves it had, how far apart they were and how many participants were lost by the final wave.
Working through Longitudinal Designs and Change Over Time in 13344? Sia is AskSia’s AI Psychology tutor — ask any 13344 Longitudinal Designs and Change Over Time question and get a clear, step-by-step explanation grounded in how 13344 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.