Lesson 9: Investigating sleep: a correlational study and Pearson's r
Design a correlational study of music before bed and sleep
SkillsDesigning an investigationCorrelation and Pearson's r
Worked example
A student wants to investigate whether listening to music before bed is related to how long people sleep. Design the investigation: state the research design and justify it, operationalise the variables, write the hypotheses, describe the sampling, give two controls for reliability or validity, address ethics, and say how the data will be analysed.
Step 1: Design
Correlational: both variables are measured as they occur. Manipulating sleep would breach the QCAA guideline against sleep deprivation in school research, and assigning music habits for a week is impractical.
Step 2: Operationalise
Music before bed: minutes of listening to music in the 60 minutes before lights out, recorded in a diary each night and averaged over 7 nights. Sleep: hours from falling asleep to waking, recorded each morning (or by a wearable), averaged over the same 7 nights.
Step 3: Hypotheses
H₀: There is no correlation between the average minutes of music before lights out and the average hours of sleep of Year 11 students. H₁: There is a correlation between them.
Step 4: Sample
A stratified random sample of 40 Year 11 students (boys and girls in proportion, chosen at random from each), so the sample represents the cohort and is large enough to detect a moderate correlation.
Step 5: Controls
The same diary and written instructions for everyone; data from the same 7 school nights; averaging over a week to reduce night-to-night variation; diaries filled in each morning, not from memory at the end.
Step 6: Ethics
Informed consent (written, with a parent or carer's written permission for students under 16), voluntary participation and the right to withdraw, coded diaries for confidentiality, normal sleep only, and a debriefing.
Step 7: Analysis
A scatterplot with a trendline; Pearson's for direction and strength; its -value compared with 0.05; a conclusion that avoids claiming cause.
Answer
A correlational study of 40 stratified randomly sampled Year 11 students, measuring average minutes of music in the hour before lights out and average hours of sleep over 7 nights by diary; H₀ no correlation, H₁ a correlation; standardised diaries and weekly averages; consent, withdrawal, confidentiality and debriefing; analysed with a scatterplot, Pearson's and its -value.
Try a similar one
A fresh question of the same kind, checked as you answer, with hints and a full solution.
Method
Choose the design and justify it.
Operationalise both variables precisely.
Write H₀ and H₁ with the operationalised variables and population.
Sampling, controls, ethics, analysis.
Watch out for
Naming an IV and DV in a correlational design. There are two co-variables.
Planning to deprive participants of sleep.