Personal Investigation 1 - A Quasi-experiment On Age And Sleep Flashcards
What factors do you think will affect sleep quality in adults
- Aging circadian rhythms (24 hour cycles),
- insufficient exposure to daylight,
- body secretes less melatonin (hormone to help you fall asleep)
- more daytime napping
What factors do you think will affect sleep quality in teensagers
- using phone before bed
- puberty
- secreting more melatonin than adults
What was the operationalised IV and DV in this investigation (methodology)
- The IV is the participant’s age; whether they are aged 20 years and under (students) or 21 years and over (teachers).
- The DV is the sleep quality as measured on an adapted version of the sleep quality scale (SQS) where higher scores (out of 30) indicate poorer sleep quality.
Describe the sample (methodology)
- There will be 20 participants (10 participants aged 20 years and under and 10 participants aged 21 years and over).
- The mean age of participants aged 20 years and under is 16 years and 4 months of age.
- And the mean age of participants aged 21 years and over is 45 years and 6 months of age.
- All of the participants in condition one (20 years and under) are all psychology students in their first year of studying A Level psychology and so are an opportunity sample.
- All of the participants in condition two (21 years and over) are teachers at the same sixth form college and so also are an opportunity sample.
- The sample will be drawn from a target population of all psychology students and staff studying at the same sixth form college in Hampshire.
Explain how you used a quasi-experiment in your research
- My research was not a ‘true’ experiment as I did NOT deliberately manipulated the IV - age (student/teacher) is a naturally occurring variable.
- My sample were NOT randomly allocated to conditions either as they were either a teacher or student and that was a difference that already existed.
Strengths of quasi-experiments
1) It allowed my research to happen as my IV of age can’t be manipulated for practical or ethical reasons. This meant I could investigate a variable like age.
OR
2) It allowed me to investigate ‘real’ problems, such as the effects of age on sleep quality, which can help more teachers and students understand sleep better.
Weaknesses of quasi-experiments
1) I cannot demonstrate causal relationships because my IV of age (student/teacher) wasn’t manipulated directly, so I cannot be sure that the a change in age caused the different sleep quality scale scores.
OR
2) There was a threat to internal validity due to there being less control over extraneous variables that could be the reason for the DV (sleep quality scores) rather than the IV (students/teachers) as the IV was not manipulated.
Describe the procedure
- Participants received an online questionnaire (SQS) via email.
- SQS had 10 questions assessing sleep quality.
- Responses were scored from 0-3 (higher scores = poorer sleep), so overall produced a score out of 30
—> Rarely (0-3 times a month) = 0
—> Sometimes (1-2 times a week) = 1
—> Often (3-5 times a week) = 2
—> Almost always (6-7 times a week) = 3 - Participants returned the completed questionnaire via email.
- A debrief was sent afterward
Explain how you operationalised the IV in your quasi-experiment
The IV was the participant’s age; whether they are aged 20 years and under (students) or 21 years and over (teachers).
Explain how you operationalised the DV in your quasi-experiment.
The DV was the sleep quality as measured on an adapted version of the sleep quality scale (SQS) where higher scores (out of 30) indicate poorer sleep quality.
Explain one possible problem with how the independent variable was operationalised
In the student condition (20 years and under) there was a smaller age range of 2 years whereas in the teacher condition (21 years and over) the age range was 25 years.
Explain one possible problem with how the independent variable was operationalised
By only collecting quantitive data to get the sleep quality score it is not as in-depth as qualitative data s their answer has been restricted to a fixed response
State the aim of you quasi-experiment
To investigate if sleep quality as measured on an adapted version of the sleep quality scale (out of 30) varies between students (aged 20 years and under) and teachers (aged 21 years and over)
Identify whether you chose a directional or non-directional alternative hypothesis
Directional
Explain the choice of hypothesis
I chose a directional hypothesis as previous research suggested that sleep quality varies by age and that sleep quality is particularly worse in adults (teachers) then in teens (students)
Write an alternative hypothesis for your quasi-experiment
There will be a significant increase in sleep quality as measured on the adapted version of a sleep quality scale (out of 30) for students (aged 20 and under) than teachers (aged 21 and over)
Write a null hypothesis for your quasi-experiment
There will be no significant difference in sleep quality as measured by the adapted version of the sleep quality scale (out of 30) between teachers (aged 21 and over) than students (aged 20 and under)
Explain one extraneous variable that you considered in your quasi-experiment
The presence of a research whilst completing the SQS could have influenced the results as my presence may have made them answer differently to if they were alone
Explain how you managed this extraneous variable in your research
So, I carried out the quasi experiment online so that l was not present at the time of completing the sleep quality questionnaire and I was not a distraction to my participants.
Explain how one confounding variable could have influenced your quasi-experiment
Some of my participants had a psychology test the day before completing the sleep quality scale so the stress of the exam may have impacted the score that they got for the dependent variable.
Outline the location of the research in your quasi-experiment.
My study was conducted online because participants were emailed a sleep quality scale in the form of a questionnaire. They were asked to complete this and send it back to me via e-mail over the Internet.
Outline one strength of conducting research on-line in your quasi-experiment
In my study, data was collected electronically by asking people 2 complete the sleep quality scale and e-mail it back to me. This was easier for me to analyse than doing it by hand.
Outline one weakness of conducting research online in your quasi-experiment
- Research online posed ethical issues in my study with informed consent.
- Although I gave my participants a consent form to read before completing the sleep quality scale, I can only assume that my participants understood what the study involved and fully agreed to take part.
Explain the experimental design in your quasi-experiment.
We used an independent groups design as different participants took part in either the student condition or teacher condition. Therefore my participants completed only one condition based on their age.
Outline one strength of the experimental design used in your quasi-experiment
There were less demand characteristics As participants only took part in either the student or teacher condition and did not experience both so they were not aware that I was interested in measuring differences in sleep quality based on age.
Outline one weakness of the experimental design used in your quasi-experiment
There were more participants variables which may have impacted my results as some of the participants in the student condition had a psychology test the next day, which may have affected the sleep quality they then gained
Explain one ethical issue which you considered in your quasi-experiment.
One issue was deception because I didn’t tell my participants that the aim of the study was about differences in sleep quality based on age, instead I just asked them to complete the sleep quality scale.
Or
One ethical issue was risk of harm because my participants could be embarrassed about recording the quality of sleep that they have as I was going to be analysing their data.
Explain how you managed the ethical issue identified previously
I managed deception by giving the students and teachers are full debrief after the study explaining how we hid the aim of the study otherwise they may give a different sleep quality score.
Or
I managed risk of harm by telling the participants that they had the right to withdraw in the consent form that they signed. This meant that they could leave my study at any point if they felt embarrassed about being asked about their sleep quality
Identify the target population in your quasi-experiment.
All psychology students and staff studying at the same sixth form college in Hampshire
Explain how your sample is representative of the target population in your quasi-experiment.
We chose 10 students and 10 teachers from the same sixth form college. This means the results on the sleep quality scale for both ages can be generalised from the 20 participants in the sample to other psychology students and staff at the sixth form college.
Identify the sampling frame used in your quasi-experiment
Students from one class of L6th psychology students. Teachers from the humanities faculty
Identify the sampling technique used in your quasi-experiment.
Opportunity (sampling)
Explain one strength of the sampling technique used in your quasi-experiment.
It was easier for me to collect participants in comparison to other sampling techniques like self-selected as participants were selected at my convenience. I didn’t have to generate an advert and wait for people to reply to my study on sleep quality and age, saving me time and costs.
Explain one weakness of the sampling technique used in your quasi-experiment
I ended up with a biased sample, because the selection was students who were my friends in the Loth and humanities teachers that teach me, so my sample may share many attributes and is not very diverse/representative of my target population.
Explain how you used primary sources in your quasi-experiment.
I collected information on the participants’ sleep quality first-hand as I emailed my sample the questionnaire on sleep myself and didn’t rely on other researchers to collect the data.
Outline one strength of primary sources collected in your quasi-experiment
I was able to control the format of exactly how my data on sleep quality was collected. I adapted the sleep quality scale questionnaire so that it related to the aims of the research on age and sleep
Outline one weakness of primary sources collected in your quasi-experiment
I was concerned that I might get socially desirable answers from my sample as it is more socially acceptable to say you have had a better sleep quality than you actually have, so data collected may lack validity.
Explain how you collected quantitative data in your quasi-experiment
I collected numerical data using the sleep quality scale. I did this by asking 10 questions and each one was scored from 0-3, giving a total score of 30.
Explain one strength of collecting quantitative data in your experiment
My data was easier to analyse than qualitative data because it is numerical and statical tests could be used.
Or
My data was easier to collect from a large group of participants than qualitative data. I collected 20 sleep quality scale scores but it would be more time consuming to collect in-depth responses.
Explain one weakness of collecting quantitive data in your experiment
My data lost the ‘human’ level of behaviour as I asked them to agree on a scale to a statement about their sleep quality, meaning they don’t have the opportunity to give an in-depth response.
Or
My data offered a very shallow view of sleep quality because I summarised thew quality of their sleep into a score out of 30 and it wil not have the same insight as qualitative data would have.
Identify the level of measurement in your quasi-experiment
Interval
●Calculate the mean, median and mode for each condition (include workings).
Mean
Students: 10+9+12+8+4+3+6+4+11+6 = 73. 73/10 = 7.3.
Teachers: 20+8+15+9+12+17+16+8+18+14 = 137. 137/10 = 13.7
Median
Students: 3 4 4 6 6 8 9 10 11 12. 6+8= 14. 14/2 = 7.
Teachers: 8 8 9 12 14 15 16 17 18 20. 12+14 = 13.
Mode
Students: 3(1) 4(2) 6(2) 8(1) 9(1) 10(1) 11(1) 12(1). Mode is 4 and 6.
Teachers: 8(2) 9(1) 12(1) 14(1) 15(1) 16(1) 17(1) 18(1) 20(1). Mode is 8
Outline one strength of each measure of central tendency in your quasi-experiment (mean)
I chose the mean as it meant I could do further statistical analysis by calculating the standard deviation using the sleep quality scale. This then allowed me to see the spread of my data also.
Or
I chose the mean as I knew it could always be found when using ordinal or above level data, which I did in my study as I collected interval data with a sleep quality scale score out of 30.
Outline one strength of each measure of central tendency in your quasi-experiment (median)
I chose the mode as it wouldn’t generate a ‘nonsense’ value as the value has definitely occurred in the data set. For teachers the mode was 8 as 2 teachers scored 8 on the sleep quality scale whereas 13.7 (the mean) is a nonsense value.
Or
It is not influenced by anomalous results. For example, Participant 11 scored 20 and this deviates from the other teachers. The mode will not be artificially inflated by this result as the most common value is selected instead.
Outline one strength of each measure of central tendency in your quasi-experiment (mode)
It is not influenced by anomalous results. For example, Participant 11 scored 20 and this deviates from the other teachers. The median will not be artificially inflated by this result as the middle value is selected instead.
Or
It can always be found when using ordinal or above level data. I collected interval data as it was a score on a sleep quality scale but if I used nominal data like how is your sleep quality? (poor/good) then I wouldn’t be able to use the median.
Outline one weakness of each measure of central tendency in your quasi-experiment (mean)
It was influenced by the anomalous results in my data. For example, Participant 11 scored 20 and this deviates from the other teachers. The mean was artificially inflated by this result as all data points are included.
Or
Calculating the mean produced a ‘nonsense’ value not in the original data set. The mean sleep quality for the teachers was 13.7 and students was 7.3 but this is a nonsense value as no participant got that score.
Outline one weakness of each measure of central tendency in your quasi-experiment (median)
In my study there was more than one mode - for students the mode was 4 and 6 and I would prefer one number to summarise the data.
Or
I had a skewed distribution in my student condition as the mode was 4 and 6 but the median was 7 so my mode did not display what is occurring in the centre of the data set.
Outline one weakness of each measure of central tendency in your quasi-experiment (mode)
It was not useful in my further statistical analysis as I calculated the standard deviation but the median score on the sleep quality scale wasn’t used in that calculation.
Or
Calculating the mean produced a ‘nonsense’ value not in the original data set. The median sleep quality for the students was 7 but this is a nonsense value as no participant got that score.
Explain which measure of central tendency is the most appropriate for your quasi-experiment
The mean is the most appropriate because it is necessary for me to carry out further statistical analysis such as standard deviation on the sleep quality scale, whereas the mode would not allow me to calculate SD from it.
Or
The mode is the most appropriate because when I calculated it, it was mot influenced by outliers like people who had a sleep quality, whereas if I had chosen to calculate the mean value it would’ve been influenced
Calculate the range for condition 1 - aged 20 years and under (students) (include workings)
12 - 3 = 9
Calculate the range for condition 2 - aged 21 years and over (teachers) (include workings)
20 - 8 =12
Outline one strength of calculating the range in your quasi-experiment.
The range for the sleep quality scale scores was easier for me to calculate than the standard deviation, so it was less time consuming.
OR
It was appropriate as I collected frequency data as my sample of teachers and students gave me a score for their sleep quality (out off 30).
Outline one weakness of calculating the range in your quasi-experiment
The range was influenced by anomalous results as Participant 11 scored 20 and this is more than the other teachers; this is a problem because only the highest and lowest scores are considered in the calculation.
OR
My range would fail to consider the distribution of scores around the mean for their sleep quality scale score, so I won’t know if the scores are close to the mean or not.
What is the formula for calculating the standard deviation
Calculate the standard deviation for condition 1: aged 20 years and under (students) (include workings)
Calculate the standard deviation for condition 2: aged 21 years and over (teachers) (include workings)
Outline one conclusion that can be drawn from these 2 standard deviations (students = 3.164, teachers = 4.3)
There is a larger variation around the mean for teachers as their SD was (4.3) whereas there was a smaller variation around the mean for students as their SD was (3.164).
Outline one strength of calculating the standard deviation in your quasi-experiment
When I calculated SD it gave me a more representative measure of dispersion unlike the range, because all of the sleep quality scale scores are all reflected every score in the data set (unlike the range). It gave me an indication of how close most of the sleep quality scale scores were to the mean.
Outline one weakness of calculating the standard deviation in your quasi-experiment
It was more time consuming when I calculated the SD compared to the range as I had to use a complicated formula for SD but with the range it would be quicker to minus the highest score from the lowest.
Or
Not all of the sleep quality scale scores will be within one standard deviation, so it can be misleading when it comes to anomalies, like participant 11 who scored the highest score for teachers, which was 20.
Explain why calculating a standard deviation is better than calculating the range for your quasi-experiment
When I calculated SD it gave me a more representative measure of dispersion unlike the range, because all of the sleep quality scale scores are all reflected every score in the data set (unlike the range).
Explain how the data from your quasi-experiment is plotted on a bar chart
Age was plotted along the x-axis (students and teachers). The mean score on the sleep quality scale was plotted on the y-axis. The mean score on the sleep quality scale for teachers and students was presented as 2 separate bars.
Explain why a bar chart is the appropriate graphical representation for your data
The bar chart represents frequencies of noncontinuous (discrete) data. The frequencies was for the mean score on the sleep quality scale and the discrete data was the two ages (students and teachers).
Draw a bar chart for the data in your quasi-experiment
Outline one conclusion from the bar chart you have drawn
Teachers scored higher on the sleep quality scale as their mean was 13.7 and students scored lower as their mean was 7.3.
identify and fully justify the inferential statistic you use to analyse the data in your quasi experiment.
- Independent measures design as my sample were either teachers or students
- My data was at least ordinal as my sample gave a score out of 30 on the sleep quality scale
- I was testing for a difference in sleep quality between students and
teachers
What was our observed (or calculated) and critical value
- observed (or calculated) value = 13
- critical value = 27
Explain whether or not you found a significant result in your quasi-experiment.
Because the observed value of 13 is lower than the critical value of 27 so the results are significant and my null hypothesis is rejected.
Explain one issue of reliability in your quasi-experiment + explain how you dealt with the issue of reliability identified previously
One issue could be the time-of-day participants filled out my sleep quality scale as some may say they had worse sleep in the morning than in the afternoon meaning it is not the same.
I dealt with this by emailing out the sleep quality scale questionnaire at 8am and asked my sample to complete it by 12pm that same day so everyone completed it in the morning.
Explain one issue of validity in your quasi-experiment + Explain how you dealt with the issue of validity identified previously
The presence of me, the researcher, whilst completing the sleep quality scale could have influenced the results as I may make them answer differently to when they are alone.
So, I carried out the quasi experiment online so that I was not present at the time of completing the sleep quality questionnaire and I was not a distraction to my participants.
Suggest 2 ways in which your quasi-experiment could have been improved
I could make my study more ethical by ensuring confidentiality as I collected the names of everyone on my sample. I would do this by collecting their sleep quality scale results but not making a note of their names. This would make the study more valid as my participants will be more likely to be truthful about their sleep quality if they cannot be identified.
I could use a random sampling technique instead of opportunity sampling. I would do this by collecting all of the names of L6th psychology students and putting them in a hat to select 10 and then do the same for the teachers in the humanities faculty. This would reduce researcher bias as I would not be selecting the sample, everyone has an equal chance of being selected.
