Personal Investigation (age+sleep) Flashcards
What was the IV for our experiment?
Age, whether they were a young adult (16-24) or an adult (25+).
What was the DV for our experiment?
Sleep measured by a score on the Pittsburgh Quality of Sleep Index (PQSI).
Give the alternative operationalised hypothesis.
There will be an increase in the score on the PQSI in adolescent participants (ps) aged 16-24 compared to adult ps aged 25+.
Is our alternative hypothesis directional or non-directional?
Directional.
Why did we use directional/non-directional hypothesis?
Research on sleep clearly states that adolescents and young adults need more sleep than adults. And, due to the phase delay in sleep patterns of adolescents, which is mismatched with school/work, teens sleep patterns don’t match their circadian rythmes, leading to poor sleep quality. So we chose directional based on this as we would expect higher PQSI scores in teens than adults.
State our null hypothesis.
There will be no significant difference in the score on the PQSI in adolescent ps aged 16-24 compared to adult ps aged 25+. Any difference will be due to chance.
What are the main characteristics of our sample?
20 people.
Mix of genders.
Aged 16-24 and 25+.
All from the East Midlands.
What sampling technique did we use and why?
Opportunity.
+It is more convenient than stratified as it is less time consuming to gather 16-24 and 25+ ps due to it not involving any calculations or the whole list of ps from the target population.
+As the p is being approached by us, there are less issues surrounding consent compared to random sampling where the 16-24 and 25+ ps would be selected and then could say no when approached by us.
What are the limitations of our sampling technique?
-Just because people are available to us to take part in our study, does not mean they are willing to.
-Our sample will be biased as as we are selected ps, we are more likely to go up to people we know or who look friendly.
Give the procedures to how we ‘will’ conduct our study.
1- We will approach Bilb college students in canteen and ask if they are willing to take part in our study. For the adults, we will approach Bilb college staff and people around Nottingham area and ask if they would take part.
2- We will give ps from the local area a brief that outline the study on age/sleep which gives details of the PQSI that they have to complete and if the are happy then they sign our consent form.
3- After, they will be asked to tick the 25+ or 16-24 box on our table.
4- We will share the link to the PQSI test and ask them to fill it in. They will give us their score on the PQSI and we will record on our results table.
5- They will be debriefed, hypotheses will be outlined, as well as true purpose of the study and give them a chance to ask qs or withdraw. If they are happy then they sign the consent form to finish the study.
Which descriptive statistics can describe our data and why?
The mean as the DV is measured as ‘score on PQSI’ which is a score from 0-21 with a set scale and true 0, making our data ratio. So the mean is the most appropriate measure of central tendency.
The standard deviation is the most appropriate measure of dispersion as we can calculate the mean PQSI score for teens/adults. As the mean is required, it gives a more accurate spread of data.
What graphical representation did we use and why?
Bar chart as the graph is displaying the mean scores on the PQSI of teens and adults, this data is discrete and therefore a bar chart is the most appropriate way to display this data.
What inferential statistic did we use and why?
Mann Whitney U. Most appropriate for our data as the experiment is looking for the difference in scores on the PQSI for teens compared to adults. Experimental design is independent groups as ps only complete PQSI once and as IV is age, ps can only belong to one condition (teen/adult). The level of measurement is at least ordinal as the DV is measured as the score on PQSI, which is a score from 0-21.
What was the mean and standard deviation of the 16-24 year olds?
Mean - 10.2
Standard deviation - 3.22.
What was the mean and standard deviation of the 25+ year olds?
Mean - 5.5.
Standard deviation - 2.42.
What was the observed value?
13.
What was the critical value?
27.
What conclusions can we draw?
For a Mann Whitney U test, as the calculated value was 13, which is lower than the critical value of 27, the findings were significant at p=0.05, therefore the alternative hypothesis was accepted. This shows that adolescents had a higher score on the PQSI than adults, therefore have worse sleep quality.
What issues of reliability did we face and how did we deal with this?
As a group we did not conduct the research together. This means the way we delivered the instructions to the 16-24 and 25+ year olds may not have been consistent between each researcher. this would affect internal reliability. To deal with this we created standardised instructions on how the PQSI was to be completed. This improved internal reliability.
Another issue was not being able to replicate due to time constraints so we couldn’t assess external reliability. To deal, we would do test-retest method. After leaving a time gap of a week, we would approach the same ps and ask them to repeat the PQSI.
How did we establish our study was reliable?
Through clear operationalisation of IV/DV.
IV - age; 16-24 or 25+.
DV - sleep; sleep quality measured through PQSI.
We also used the split half method. We would split ps answers on the PQSI in half (odd and even) and compare scores on each half. If scores are consistent, then the internal reliability of the questionnaire in measuring sleep quality is high.
What issues of validity did we face and how did we deal with them?
Qs on PQSI may have been difficult to understand, especially for teen group (16-24). This means they may have answered the qs accurately, affecting internal validity of the results. We dealt with this by getting ps to complete the PQSI in our presence so they could ask qs.
Also during the piolet study, we saw ps being distracted in canteen (discussing answers etc). To deal with this we asked p to only complete PQSI in our presence and no one else to avoid distractions. We took each person to a quiet area to avoid this.
How did we establish our research to be valid?
Content validity - ask a sleep expert if the PQSI is an accurate measure of sleep quality. If they agree, then the research has good content validity.
Operationalisation - clear operationalisation of age (16-24 or 25+) and sleep (PQSI 0-21) increases internal validity of our research.
Concurrent validity - compare scores on PQSI to another established sleep quality measure. If both measures give similar results on sleep quality then our research could be said to have good concurrent validity.
What were 2 ethical issues that concerned us before the research?
Lack of confidentiality - eg, collecting names, ages and other personal details.
Lack of protection from harm - qs on questionnaire can be personal - eg, asking about health issues which may make the ps feel uncomfortable.
How did we deal with these ethical issues?
Confidentiality - we will only gather data that is necessary for the study like age and ps will be kept anonymous by using numbers instead.
Protection from harm - ps will be briefed about the sorts of qs they can be asked as well as debriefed if they want to remove their data.
Give 2 ways that we could have improved our research.
More varied/diverse sample of 16-24 and 25+. All were from small area of East Midlands. Do this by conducting study online to reach a wider and more diverse sample. Increase population validity of study, meaning sample is more likely to be representative of the effect of age on sleep quality.
I would improve measure of sleep chosen. PQSI is questionnaire designed to measure sleep quality, but some of qs were difficult to answer accurately. ‘Over past month, have you woken to go to toilet?’. To improve, i would measure sleep quality in a sleep lab, where sleep can be accurately measured using EEG scans. This would improve internal validity.
