personal investigation - quasi experiment Flashcards
aim?
The aim was to see if the quality of sleep was different between young people and older people
hypothesis ?
There will be a difference in the score on the sleep quality questionnaire between participants aged 17 to 25, and participants aged over 50.
null hypothesis ?
There will be no difference in the score on the sleep quality questionnaire between participants aged 17 to 25, and participants aged over 50.
what hypothesis was used and why?
A non-directional hypothesis was used because, previous research has shown older adults have poorer quality sleep due to changes in hormones and aging, recent research has indicated that young adults engage in poor sleep hygiene (inconsistent sleep schedules, excessive screen time before bed, caffeine intake close to bedtime, and not having a relaxing bedtime routine) which affect sleep quality. Therefore previous research is inconclusive.
indépendant variable ?
whether the age of the participant - whether they were young adults (17 to 25) or older adults (over 50).
age cannot be manipulated which is why this quasi is not a true experiment
dependant variable ?
DV was sleep quality. We used the Pittsburgh Sleep Quality Index (PSQI) to measure sleep quality which is a questionnaire with multiple choice answers. The maximum score is 42 which would indicate very poor sleep quality, the lowest score would be 0 (amazing night’s sleep).
why was the DV sleep quality ?
We chose to operationalise sleep quality using the PSQI because it was a validated measure of sleep quality, and using a questionnaire would be an efficient way to measure sleep quality.
strength of quasi ?
- more practical and feasible than a true experiment. we were available to study age related differences without assigning PT to different age groups which would been impossible
- real world relevance since PT completed a questionnaire on their sleep quality which gave us ecological validity and reflect real world sleep behaviours than a tight controlled lab experiment
weakness of quasi ?
- IV was not manipulated but PT were either young or old we cannot assume cause and effect. if we found a difference in sleep quality between the ages this may be due to other differences in the PT such as illness
- factors such as environmental conditions, diet to psychological well being can vary across age groups which makes it challenges to isolate the effect of age alone
location of research ?
online - using self report questionnaire on a google forms questionnaire this was emailed to PT
strengths of location of research ?
- convince and accessibility - PT could complete the questionnaire on sleep quality at ether own convenience making it easier to reach large and diverse sample
- cost effective - as we sent the questionnaire which eliminated costs associated with printing
- large sample size improving generalisability
weakness of location of research ?
- limited control over responses - without supervision our PT may have rushed through the questionnaire which may have provided careless answers or misunderstood questions
- sampling bias - certain groups may be underrepresented limiting the generalisability
experimental design ?
independent measures design as our PT were either young adults to older adults
advantage of experimental design ?
- as we were looking at age groups this was the most feasible
disadvantage of experimental design ?
- results could be affected by PT variables lowering external validity
- need to find more PT to get the same number of data then a repeated measure design we had an overall sample group of 76 however this was reduced to 34 in the younger group and 42 in the older group
extraneous variables ?
- social desirability bias = PT may have been embarrassed by their sleep patterns and may not wanted to admit any issues so affecting the validity
- researcher bias - We predicted there would be a difference between the age groups in terms of sleep quality and therefore out behaviour (intentionally or unintentionally) could have affected the way the participants answered the questions.
how to deal with extraneous variables ?
- social desirability bias = we chose to use an online questionnaire to measure sleep quality rather than an interview or a questionnaire where the researcher was present. Also we only collected email addresses if the participant wanted to know the results of the questionnaire but otherwise they remained anonymous. These mean that the participants could be honest without feeling judged or embarrassed
- researcher bias = by choosing to use an online questionnaire to collect our data means that we would be unable to influence the results of their answers in the questionnaire on sleep quality.
sample ?
-opportunity sample, researchers emailed friends and family
- 76 PT
- 34 PT for young adult
- 42 PT older adult
strength of sample ?
we were able to gather our sample from our sampling frame (people between 17 and 25 and over 50) very quickly so it was less time consuming and also more convenient.If we used another method such as random sampling we would have had to get hold of all the names and email addresses of people who were in the age groups of interest which would have taken time and would be more difficult to manage. So opportunity sampling was far more convenient.
weakness of sample ?
our sample lacks population validity as it’s a small portion of the target population which means the sample is not as representative making it harder to generalise findings to the wider population. As they were people who were known to the researchers who were psychology students - we are only reaching a small proportion of the population, and many of the younger age group would be A Level college students who may not represent the whole population of 17 to 25 year olds
procedure of quasi ?
- questionnaire was used to measure sleep quality based on the Pittsburg sleep quality index, asked on google forms and sent to PT to complete online
- The questions started by asking “During the past month how often have you had trouble sleeping because….” , and this was followed by 14 variables such as “couldn’t get to sleep within 30 minutes”, “feel too hot”, “had to go to the toilet”. They then had a multiple choices of “not during the past month”, “less than once a week”,”once or twice a week”, “More than three times a week”. There was a question on which age group the participants were (17 to 25 or over 50)
- Researchers emailed the questionnaire to people who they knew in the right age groups using an opportunity sample
- The questionnaire included a consent form that explained the full aims of the experiment, that the questionnaire was anonymous and confidential, that it would take no longer than 5 minutes to complete, they could withdraw at anytime and that it was for a Psychology A Level. If they consented they went on to complete questionnaire and there was a debrief form at the end which gave them the opportunity to withdraw again and leave their email address only if they wanted to know their score.
- The questionnaire was scored out of 42, where a low score was good sleep quality and a high score was poor sleep quality.
descriptive statistics justification and findings ?
We calculated the mean and the standard deviation for the two age groups of the experiment .
Justification: The mean gave us one value for each age group (17 to 25, and over 50), that made the results easier to compare. Also the mean is the most sensitive measure of central tendency as it takes all the data (in this case the score on the sleep quality questionnaire) into account. We also calculated the standard deviation for the 3 conditions (see above).
Justification: The SD was used as it gave us one value for each age group that made the results easier to compare. It is the most sensitive measure of dispersion as it takes all the data (in this case the score on the sleep quality questionnaire) into account
graphical representation and justification ?
We divided the sleep quality scores into groups of scores representing very good, good, fair, poor, very poor sleep quality. We calculated the percentage of participants who scored in each category for each age group and drew a composite bar chart.
This was approriate because we are comparing categories (of sleep qualities) and a bar chart will allow us to compare these different sleep quality categories for each age group.
inferential statistics justification ?
We used the Mann Whitney U test to analyse the data.
Justification
We are looking for a difference between the sleep quality scores between young adults (17 to 25) and older adults (over 50).
The experimental design was independent measures (unrelated) because the participants were either 17 to 25 years old, or over 50.
The data collected was a score on the sleep awareness questionnaire. This is interval data as it is a measure on a scale that has equal intervals, but does not have a true o. Interval data can be ranked and treated as ordinal data.
inferential statistics findings ?
We chose to use the P≤0.05 level of significance as this is the minimum level of significance accepted in psychology and represents a 5% chance that our results on context dependent forgetting could be a fluke. As our research was on memory and not something crucial that we need to be really sure of (like treatments) then this level of chance was seen as acceptable.
Our hypothesis was a directional hypothesis as we predicted more words would be recalled when participants learned and recalled in the same context than if they didn’t.
The observed value of U was 450.5.
As the observed value of U was less than the critical table value the test at the P≤0.05 level of significance. Therefore the test was significant and we rejected our null hypothesis and accepted our research hypothesis which stated “There will be a difference in the score on the sleep quality questionnaire between participants aged 17 to 25, and participants aged over 50.”
summary of findings ?
the mean sleep score for the 17 to 25 age group was 16 compared to the mean sleep score of over 50 which was 12. This means on average the younger age group have poorer sleep quality than older adults.
The bar chart also supported this as a greater percentage of 17 to 25 year olds had poor sleep quality compared to over 25s.
The mann whitney U test result was significant at the P≤0.05 level of significance. Therefore the test was significant and we rejected our null hypothesis and accepted our research hypothesis which stated “There will be a difference in the score on the sleep quality questionnaire between participants aged 17 to 25, and participants aged over 50.”
summary of conclusions ?
as the inferential test (Mann Whitney U) was significant this shows that there is a difference in the sleep quality between younger and older adults. Using the descriptive statistics we can say that younger people have worse quality of sleep than older adults.
issues with reliability?
f the questionnaire on sleep quality was not standardized, poorly worded, or lacks clear response options, it can lead to inconsistent answers.
Self-Report Bias
People may not accurately recall or honestly report their sleep quality due to embarrassment or memory. This will result in inconsistent results
how to deal with issues of reliability?
to deal with questionnaire design, we used a standardized and validated sleep questionnaire (e.g., Pittsburgh Sleep Quality Index). We also made sure our instructions to participants were clear and written at the beginning of the google forms so everyone had consistent experience. This would improve the internal reliability of the questionnaire
To deal with Self report bias we ensured anonymity to reduce social desirability bias, making participants more likely to report their true experiences of sleep.This improved the external reliability
how we assessed the external reliability of the experiment ?
in order to test the external reliability we could repeat the quasi-experiment with the same group of participants and ask them to compete the sleep quality questionnaire at a later date - maybe a month later - this is called test retest reliability. If the correlation coefficient was greater than 0.8 there would be good test retest reliability.
issues with validity ?
social desirability bias - Our participants may have been embarrassed by their sleep patterns and may not have wanted to admit to themselves or a researcher any issues with the quality of sleep - so they could not be totally honest in their answers on the sleep quality questionnaire. This would affect the internal validity.
Researcher bias - We predicted there would be a difference between the age groups in terms of sleep quality and therefore out behaviour (intentionally or unintentionally) could have affected the way the participants answered the questions. Again this would affect internal validity
how we dealt with issues of validity
Dealing with social desirability bias: we chose to use an online questionnaire to measure sleep quality rather than an interview or a questionnaire where the researcher was present. Also we only collected email addresses if the participant wanted to know the results of the questionnaire but otherwise they remained anonymous. These mean that the participants could be honest without feeling judged or embarrassed.
Dealing with researcher bias: - by choosing to use an online questionnaire to collect our data means that we would be unable to influence the results of their answers in the questionnaire on sleep quality.
how we assessed the validity of the experiment ?
to assess validity we could use concurrent validity. We could find another established measure of sleep quality that psychologists have used - for example and interview and assessment on sleep quality with a clinical psychologist. We could get a group of participants to complete both our sleep quality questionnaire and the obtain a clinical psychologist rating on sleep quality.
We could then conduct a correlational analysis to see if there was a positive relationship between the scores on both assessments. A correlation coefficient of 0.8 or more would indicate good concurrent validity.
ethical issues ?
Psychological harm. It is possible some participants may have felt mildly anxious after completing the questionnaire as it may have highlighted any potential sleep difficulties they had.
Confidentiality: Participants may not have wanted others to be aware of their sleep difficulties and their score on the sleep awareness questionnaire. However they also had a right to know their score on the sleep quality questionnaire which we couldn’t give if we did not have email addresses
Consent and deception was not an issue as the participants were told the aims of the study, what was involved and how the results would be used in a consent form at the beginning of the questionnaire - no deception was used. They were also given the right to withdraw at any point.
how we dealt with any ethical issues
Dealing with psychological harm: in the consent form that they were given beforehand they were clearly told what the study was about and that they could withdraw at anytime. After they completed the questionnaire on sleep quality there was a debrief section at the end of the questionnaire that suggested if they were worried about their sleep they should contact their GP. It also gave the names of psychology teachers that they could get in touch with.
Dealing with confidentiality: The questionnaire was anonymous and we did not ask for email addresses. However in the debrief form at the end of the sleep questionnaire, participants were given the option of giving their email address if they wanted to know their sleep quality score. They were assured if they did no names would be used in the analysis.
