Вы находитесь на странице: 1из 21

QUESTIONS AND ANSWERS

The Can You questions in chapters 1-15 of the A2 Complete Companion require extended answers which are largely just a reworking of the text, so we havent provided the answers for these questions. Answers for the questions in Chapter 16 Research methods are presented here. These are suggested answers to the questions. Often there is no single right answer. In some cases e.g. has been included in the answer but even where this is not given you may assume that there could be other, perfectly correct answers. There are prizes for anyone who reports an answer (or question) that is wrong. Write to cara.flanagan@btopenworld.com

QUESTIONS 16.1
1.

Science is empirical It is easy to claim things are true but they need to be verified through direct observation. Science is objective If observations are influenced by the beliefs of the observer they are subjective. To avoid this observers aim to be objective. Scientific research can be replicated The way to demonstrate whether the result of a research study is real or true is to repeat it exactly.

2.

For example Science is empirical Milgram created a situation where we could observe the extent to which people obeyed. Science is objective Milgram created an objective means of testing obedience e.g. the experimenter had a series of pre-determined prods, all other influences were removed. Scientific research can be replicated The details of Milgrams study were carefully recorded so others could (and did) repeat it, demonstrating the validity or not of his fndings.

3. Criticisms can be positive or negative, for example: Scientific methods in psychology are desirable as a means of producing verifiable knowledge, rather than just having peoples opinions about the causes of human behaviour. For example people often believe that watching violent films makes people more aggressive. So we need a means of testing this objectively. Science offers this method. Lack of objectivity In psychology the object of study reacts to the researcher and this leads to problems, such as experimenter bias and demand characteristics, which compromise validity.

However, similar problems apply to the hard sciences. Heisenberg (1927) argued that it is not even possible to measure subatomic particles without altering its behaviour. Scientific methods havent worked Perhaps the way to decide whether science is appropriate for psychology is to look at the results of psychological research. For example, psychological approaches to treating mental illness have had at best modest success, as in the case of drug therapies which have been shown to be only marginally better than placebos. So psychologists havent discovered means of treating mental illness effectively. This suggests that the goals of science are not always appropriate for psychology.

4. Harriet could tell her mother that it is important to be objective and use empirical methods to test the hypothesis without bias. For example you could send out questionnaires to people about how much fun they had in daily life and see if blondes reported more fun. The questionnaire would have to be designed carefully to get truthful answers. Harriet could also tell her mother that replication was important to confirm any results. So it isnt enough to do just one study testing the hypothesis. But other studies should be used to confirm the original results. For example you could conduct an observational study of blondes and non-blondes and record their behaviours using behavioural categories.

QUESTIONS 16.2
1. Peer review refers to the process by which academic equals (experts in a particular field) present critical information about research that has been conducted, in order to judge the value of a piece of research. 2. Peer review is essential to the process of producing valid scientific data because it provides a means of assessing the value of the research. For example, an expert (peer) is in a position to judge whether the methods used are unbiased and well controlled because the expert has experience in that field. If the methods are biased and/or poorly controlled then the study would lack validity. A peer reviewer can also judge the quality of the research in terms of the sample of participants used in the study. If the sample is small and/or taken from a small part of the general population then the study would lack generalisability. [110 words] 3. e.g. It is not always possible to find an expert in a particular field. This means that the person who is used may not be able to judge the value of the research accurately. They may feel that the study contributes nothing and therefore it isnt published. Anonymity may enable people to present hostile reviews because they cant be blamed or attacked. However it may also allow reviewers to give overly critical reviews because they dont wish to see certain research published scientists have rivalries just like anyone else. So anonymity might not provide valid peer review. Peer reviews tend support the status quo. For example journals prefer to publish research that confirms popular theories rather than studies that promote a new and untested view. This tends to slow the progress of science. 4.

(a) It would be desirable for his research to be peer reviewed to check whether the methods used were objective and unbiased. It is important to know if the researcher had used sloppy methods because then the findings and conclusions would lack validity. For example there might be extraneous, uncontrolled variables. It would also be desirable for the research to be peer reviewed to decide on the value of the research. For example the research should be testing an explanation or have real world application. It is also important that the research is genuine and not based on invented data. (= 99 words, 4 marks should be about 100 words) (b) The paper may have been rejected because it was simply a direct replication of a previous study. Journals tend to refuse such replications because they feel their readership are not interested in them and they prefer more eye-catching stories to help their sales, like any magazine. A second reason why the study may have been rejected is because the reviewers were suspicious about some of the details of the methods used. They may have felt there was insufficient detail of the participants and how the data was collected. They usually ask a researcher to redraft the article providing more detail. (= 100 words)

QUESTIONS 16.3
1. Lab experiment e.g. Loftus and Palmer (1974), estimating speed of vehicles in an apparent car accident. Strength = extraneous variables can be controlled other factors might have influenced the DV such as seeing the car moving. Everything but the leading question was kept constant. Weakness = the contrived nature of the study means it was not like everyday eyewitness testimony where a witness may be more emotionally aroused. Field experiment e.g. Johnson and Scott (1976), weapon focus study Strength = participants were not aware that their behaviour was being studied and therefore their observations of the apparent assailant was natural. Weakness = there may have been investigator effects when participants were being questioned afterwards if the researcher knew which condition a participant was in, for example if they knew the participant was in the pen condition the investigator might have persisted a bit longer in asking the participant to describe what they saw. Natural experiment e.g. Christianson and Hubinette (1993), the effects of anxiety on eyewitness recall studying real-life witnesses Strength = gives us the opportunity to see how real-life eyewitnesses respond to high emotional arousal, increasing the generalisability of the findings. Weakness = the IV (being more or less threatened) was not controlled (participants were not randomly assigned to groups), therefore we cant conclude that the high anxiety actually caused the better recall (a causal relationship is not demonstrated in a natural experiment. Questionnaire e.g. Rahe et al. (1970), study of life changes in men on a Naval tour of duty Strength = can easily give the questionnaire out to hundreds of people in a short space of time and get responses which can be compared between participants.
3

Weakness = people may not answer truthfully, for example they may not list all the unpleasant life events that have happened to them (such as having trouble with their boss) because they want to put themselves in a better light. Interview e.g. Yuille and Cutshall (1986) interviewed 13 people who had witnessed an armed robbery. Strength = permits the interviewer to find out unexpected information because can follow up respondents answer wit another question, for example when the respondent describes something that happened the interviewer can then focus on a particular element to get more information about the crime. Weakness = interviewer bias may mean that answers are not honest but supplying the answer that the interviewer wants to hear, so the interviewer may lead the respondent to describe the armed robbery as more or less distressing. Naturalistic observation e.g. Ainsworth (1967), observations of mothers and infants in Uganda. Strength = gives a realistic picture of what people actually do in their everyday lives, not affected by being in a contrived situation in this case could see how mothers naturally interact with their infants. Weakness = the observer may see want they want to see i.e. their observations are affected by expectations so Ainsworth may have interpreted the mother-infant interactions as more sensitive than they actually were. 2. (a) demand characteristics a demand characteristic is something in a research setting that reveals the research hypothesis, thus leading participants to behave as expected. For example, it can be argued that the scientific authority figure (experimenter) in Milgrams experiment acted as a demand characteristic because people found it harder to resist than they would other authority figures. Subsequent replications (Milgrams variations) showed that obedience rates dropped when these demand characteristics were changed. This means that the baseline study only tells us about obedience in one setting and that the findings may not be generalisable to real life. (b) pilot study Asch might have conducted a pilot study prior to his main study to see how participants would react to the setup. It could have been that participants would be quite suspicious about the behaviour of the confederates so it would not have been worth doing the full scale study. The pilot study would be important in testing the procedures. (c) social desirability bias the hassles and uplifts scale is a questionnaire, participants filling this in may not be totally honest about the hassles they have experienced because they might think such hassles make them look stupid or foolish. People prefer to look good. (d) extraneous variable in Peterson and Petersons experiment the IV was time interval and the DV was recall of nonsense syllables. Rehearsal was controlled by asking participants to count backwards but this might have acted as an extraneous variable because recall (DV) was affected by counting backwards not just the time delay (IV). (e) longitudinal study an example of a study conducted over time is the study by Hodges and Tizard of children placed in institutional care when they were very young. They were tested at various times in their development to see what effect early privation had on their emotional and social development. 3. Study A Johnson and Scott (a) Method = field experiment, technique = questionnaire/interview to get identifications from participants afterwards.
4

(b) Hypothesis = People who view a confederate holding a blood-covered knife are less likely to accurately identify the confederates face than people who view a confederate holding a greasecovered pen. (c) Directional (d) Experimental design = independent groups, because if participants did both conditions the first would affect what they did in the second and they may well guess what the experiment is about. (e) Extraneous variable = it might be that participants were just more anxious or surprised in the knife condition and therefore this lead to paying less attention generally to the confederate, dealt with by having a control condition which created anxiety/surprise like confederate holding a banana. Study B Buss (a) Method = Natural experiment (IV is gender, DV is ratings on different measures.), technique = questionnaire. (b) Hypothesis = e.g. men rate chastity as more important in a mate than females. (c) Directional (d) Experimental design = independent groups because you couldnt do repeated measures with men and women. (e) Extraneous variable = answers might have been affected by social desirability bias (e.g. men might not want to admit that chastity matter to them), dealt with ensuring answers were confidential and anonymous. Study C Rahe et al. (a) Method = study using correlational analysis, Technique = questionnaire. (b) Hypothesis = There is a positive correlation between LCUs and illness record. (c) Directional (d) Experimental design = NA (e) Extraneous variable = NA Study D Kiecolt-Glaser et al. (a) Method = natural experiment, Technique = NA (b) Hypothesis = (1) Students immune system is less active (as measured by blood samples) during exams than before the exam period OR (2) Students blood samples are different during exams than before the exam period. (c) (1) Directional (2) Non-directional (d) Experimental design = repeated measures, why = it is better to compare the same person before and after to control for participant variables such as general levels of immune system functioning and stress levels. (e) Extraneous variable = exposure to infection which might affect immune system levels, dealt with by having a control group who were not doing exams but were also students and exposed to same illnesses. 4. Eyes looking straight ahead, talking to passenger, shouting at other drivers. 0. Closed question Do you think that it is acceptable to drive at 35 mph in a 30 mph area is acceptable? YES/NO? Open question What is your view of people driving 35 mph in a 30 mph area? 6. Hypothesis: Students who achieve greater success at school have a more positive attitude about school. Variables to study:
5

Success and school add up GCSE scores and calculate mean Positive attitude assess using a questionnaire to calculate a score Procedures: Design questionnaire and trial with a small group of students. For ease of scoring questions will be closed e.g. on a scale of 1 to 5 where 5 is excellent, rate each of the subjects you study. Select a sample of students using random sampling (select names from registers). Give details of what will be involved, guarantee confidentiality and gain informed consent. Ask participants to answer questionnaire, seated on their own in a room with no distractions. They can have as much time as they need. Debrief participants.

QUESTIONS 16.4
1. Validity = Refers to whether an observed effect is a genuine one. Reliability = A measure of consistency. Internal validity = the extent to which a researcher has measured what he/she intended to measure. External validity = the extent to which the findings can be applied beyond the research setting. Internal reliability = consistency within a set of scores or items. External reliability = consistency over time such that it is possible to obtain the same results on subsequent occasions when the measure is used. 2. Study (a) Factors that might affect validity = social desirability bias (mothers may not be honest because they want to appear to be good mothers), interviewer bias (interviewer may ask leading questions to obtain answers that confirm expectations). How to deal with these factors = emphasise importance of honesty, use a lie scale, ask questions in a way that avoids defensive answers, train interviewers. Study (b) Factors that might affect validity = social desirability bias (students may lie so they look like they are more hard working), students may forget to write in diary and make up the information later. How to deal with these factors = emphasise importance of honesty, devise a data capture system using mobile phone so that participants dont forget to record data. 3. (a) The psychologist has two memory tests, A and B. To check the reliability of each you need to use another group of participants. Give the participants test A and B and then a week later repeat this. Correlate scores on each test (A with A and B with B) to check reliability. (b) The psychologists should repeat some of the interviews a second time with the same interviewee and interviewer, and compare the answers to see if they are the same. 4. (a) Casually: wearing jeans, wearing trainers. Smartly: wearing a dress, wearing high heels. (b) Calculate inter-observer reliability and train observers if low. To calculate inter-observer reliability, collect scores from two or more observers e.g. how many times they observed a particular event

and then do a correlation. To train observers, watch video of behaviours and discuss operational definitions. (c) Event sampling each time a girl passes by note the length of time the man spent watching the girl, also tick items of clothing worn by the girl. (d) e.g. the checklist didnt provide observers with all kinds of clothing in order to determine whether the person is casually or smartly dressed. 5. Study (a) Sampling technique = volunteer sample Strength = easy to obtain willing participants by placing an ad on e.g. college noticeboard. Since the task is quite lengthy you need willing participants not just people you happen across. Weakness = sample may be biased because volunteers represent a particular profile. Their willingness may mean they are more alert to what is going on and better eyewitnesses. Study (b) Sampling technique = volunteer sample Strength = in each culture you would need men and women to answer so it would probably be easiest to advertise in a newspaper, and this would also means you would get a good range of age groups. Weakness = you are limited by the readership of the newspaper which may limit the generalisability of the findings. Study (c) Sampling technique = opportunity sample Strength = allows researchers to make use of all the men onboard a ship, an easy sample. Weakness = it is a limited sample of the men aboard just one naval cruiser, who have unique characteristics e.g. men in the navy might be more hardy and less affected by stress. Study (d) Sampling technique = random sample of university students Strength = can obtain an unbiased sample of the target population so that the students should be a varied and representative sample. Weakness = this kind of sampling takes time to identify all the students and arrange for them to take part. 6. A volunteer sample would enable the researcher to target a broad cross-section of people through several advertisements whereas with an opportunity sample the sample is likely to be more restricted. 7. With a random sample every member of the population has an equal chance of people selected, whereas this isnt true for a systematic sample where there is a system.

QUESTIONS 16.5
1. (a) Confidentiality of pupil IQ scores and GCSE results. Deal with it by not recording names. (b) Girls underage (teenage) so informed consent an issue. Deal with by asking both parents and girls.
7

(c) Invasion of privacy. Dealt with my gaining informed consent from parents. (d) Lack of informed consent because you couldnt tell them before they were asked to obey, deal with it by asking people afterwards and giving them the right to withdraw their data if not happy. Could also use presumptive consent. (e) Lack of informed consent as students (and their parents) shouldnt be informed beforehand that the programme was being assessed as this might affect the students performance. Deal with it by asking students and parents afterwards and explaining that both programmes were believed to be equally good (so students education not compromised, and giving them the right to withdraw their data if not happy. (f) Psychological harm and deception because some were led to believe they had done less well. They would need to be very carefully debriefed afterwards and assured that they actually did much better. (g) Deception, deal with it by debriefing students afterwards and giving them the right to withdraw their data if not happy. 1. e.g. interviewing young people with eating problems about their disorder. Pros: opportunity to gain insights into the disorder that may help people suffering from eating disorders. Cons: Interviewees may feel distressed talking about their disorder, they may be untruthful and then people get the wrong impression of people with eating disorders. 2. UK law the Animals (Scientific Procedures) Act (1986), and also BPS guidelines. 3. (Study 1) Harlows study on attachment with rhesus monkeys, ethical issues raised: Psychological harm: the monkeys were deprived of an attachment with a real mother figure which led to emotional maladjustment. Responsibility: The researchers had a duty of care to look after these animals for the rest of their lives. Numbers of animals: perhaps more used than was necessary. (Study 2) Skinners study demonstrating operant conditioning in pigeons, ethical issues raised: Psychological harm was minor in this study, the pigeons were conditioned to press a button to obtain food. There was not much stress involved and the pigeons might well unlearn the response fairly quickly. Responsibility: The animals would have to be well cared for during the study, making the environment and food as close to what they would experience in their normal lives.

QUESTIONS 16.6
1. It means that the chance (or probability) of obtaining a particular set of results is less than 5% if the null hypothesis were true. 2. 0.01 is used when testing something very important to be even more sure that the results are not due to chance, or used when replicating a study to be confident in the significance of the results. 3. Calculated value. 4. (1) Number of participants in each condition, (2) whether the hypothesis is directional or non-directional, (3) significance level selected, (4) whether the critical value has to be more or less than the observed value for significance to be shown.
8

5. (a) Ordinal (b) Interval (c) Nominal 6. The likelihood of a Type I error is 5%, a 5% chance that a null hypothesis that was true has been erroneously rejected.

QUESTIONS 16.7
1. Informed consent Make sure participants understand what will be assessed and how their data will be used. Confidentiality Numeracy scores should be confidential. 2. Problems People may give consent without truly understanding, people may be able to work out each others numeracy scores from the order of the data so the scores are therefore not confidential. 3. The ratio between the index / ring finger, and numeracy skills. 4. Testosterone (male hormone). 5. Alternative hypothesis: There is a positive correlation between the index/ring finger ratio and numeracy skills. Null hypothesis: There is no relationship between the index/ring finger ratio and numeracy skills. 6. Scattergram:

7. This is one for you.

QUESTIONS 16.8
1. Contingency table

Old Sleep more than 8 hours Sleep less than 8 hours Total 11 25 36

Young 31 33 64

Total 42 58 100

2. Should be non-directional (Chi-square always non-directional): There is an association between age and number of hours slept. Null hypothesis: There is no association between age and number of hours slept. 3. For a one-tailed test, and df = 1, the critical value would be 2.71. The observed value of 3.02 is greater and therefore significant. However for a two-tailed test, the critical value is 3.84. The observed value 3.02 is less than this and therefore not significant.

QUESTIONS 16.9
1. High bridge group Mean Median Range 7.9 8 5-10 Low bridge group 5 5 2-9

Graph should show mean and/or median not individual scores because that does not summarise the data. 3. Alternative hypothesis: Men who are more aroused (run for 2 minutes) rate women as more attractive than men who are less aroused (run for 15 seconds). Null hypothesis: There is no difference in the attractiveness ratings given by the men in the high or low arousal group. 4. Directional. 5. For a directional hypothesis (one-tailed test) the critical value of U is 33, this is less than the observed value of 40 and therefore is significant. So you would reject the null hypothesis. For a non-directional hypothesis (two-tailed test) the critical value of U is 28, this is less than the observed value of 40 and therefore is significant. So you would reject the null hypothesis.

10

6. Abstract: The aim of this study is to investigate whether physical arousal does create feelings of attraction. Participants were selected using an opportunity sample of students in the college. Male participants were in one of two conditions: either they ran on the spot for 2 minutes (high arousal) or for 15 seconds (low arousal). Afterwards they were asked to rate the attractiveness of a young woman in a video. The statistical analysis showed significance (p 0.05). Therefore we can conclude that there is a difference in the ratings given to the women in the video and that physical arousal causes men to rate women associated with the physical arousal as more attractive. (112 words perhaps a bit on the short side) Introduction: Paragraph 1: Start with general paragraph on theories of falling in love. Paragraph 2: Narrow down to Hatfield and Walsters two factor theory of love. Paragraph 3: Describe Dutton and Arons study of love on a high or low bridge. Paragraph 4: Describe the study by White et al. Explain aim of this study (to replicate the study by White et al., and give the hypothesis: Men who are more aroused (run for 2 minutes) rate women as more attractive than men who are less aroused (run for 15 seconds). (directional) Method: Details of participants and population and sampling method. Step-by-step details of what was done e.g. first 10 participants placed in high arousal condition and second 10 participants placed in low arousal condition (a systematic but relatively unbiased method of allocation to avoid participant effects). Participants tested individually. First the participant runs on the spot, then sits on own and watches video. Afterwards the participant answers a questionnaire about the video which contains a critical question about the attractiveness of the woman. Attractiveness rated on a scale of 1 to 10 where 10 is very attractive. Participants debriefed at end. Results: Descriptive statistics: mean for each group calculated (high and low arousal). Standard deviations calculated to compare whether groups show similar distribution. Bar chart used to show these results. Inferential statistics: Mann-Whitney test selected as a test of difference required for independent groups. Data is ratings, therefore ordinal. Conclusion: For a directional hypothesis (one-tailed test) the critical value of U is 33, this is less than the observed value of 40 and therefore is significant. So we can reject the null hypothesis and conclude that there is a difference in the ratings given to the women in the video and that physical arousal causes men to rate women associated with the physical arousal as more attractive. Discussion: Summarise the results. Relate to previous research which had the same findings. Consider the methodology how much can a lab study with running on the spot be generalised to real love. Implications for the two-factor theory. Suggestions for future could try with more varied sample and with different videos. Perhaps with reallife task such as watching an emotionally-arousing movie. References: as appropriate for studies mentioned e.g. Hatfield and Walster, Dutton and Aron, White et al.

11

QUESTIONS 16.10
1. At the 10% level you would require a maximum value of 100 for significance, at the 2% level you would require a maximum value of 89 for significance. 2. Alternative: People do better on a memory test in the morning than when tested in the afternoon. Null: There is no difference between memory test performance when tested in the morning or afternoon. 3. Invent your own data. 4. Choose this test because you need a test of difference (difference between morning and afternoon tests), repeated measures design (participants take part in both conditions), data is ordinal or better (test scores are interval data). 5. Use your own data. 6. Match pairs of participants in terms of a prior memory test. Then place one member of each pair in the morning group and the other member of the pair in the afternoon group. Compare partners scores. 7. Could use ABBA give participants four tests over four days: morning afternoon afternoon morning.

QUESTIONS 16.11
1. (a) Give name of organisation, researcher and address in case participants wish to contact you later. Details of procedures so participants are in a position to give informed consent, include duration and details of the procedure, as well as how data will be used. However it may be best not to reveal the precise topic in case this affects the answers given, e.g: The aims of this study are to investigate human behaviour. You will be required to ask 30 questions relating to activities you engage in and your general mood. You will not be told the full details of the study until after you have completed the questionnaire. You will be fully debriefed about this research study and the results will be shared with you. Details of pay: Participation is voluntary and there is no payment. Right to withdraw: At any time you are free to withdraw from the study. Confidentiality: You will not be asked to give your name and all information will be held confidentially. (b) IV = exercise, can be operationalised by requiring participants to take part in one of two conditions: One hour walking briskly each day or being in a control group who take no exercise. DV = happiness, can be operationalised by using a questionnaire to assess happiness, the DV is the score on the questionnaire. Extraneous variables: general patterns of exercise before the study (people who already exercise a lot might find the lack of exercise makes them unhappy and people who dont exercise might feel

12

that having to do exercise makes them unhappy), other activities during the period of the study, general health. Recruit 100 participants by advertising for people who take no exercise as part of their daily routine. Randomly allocate participants to the exercise or no exercise group. Brief each group about what is required they are not to do any kind of exercise except for the exercise group who do one hour a day of vigorous walking (make it clear what counts as exercise). The duration of the study will be 30 days. At the beginning of the study and at the end of the study participants should be given the happiness questionnaire. Participants should be given information about the study in advance and, based on this information, asked to provide informed consent. Kind of questions e.g. closed (How many minutes per day do you spend speaking with your mother/father? Rate your closeness to your different family members) and open (What do you like best about your family? Describe your relationship with your mother/father/other family members). Sample: Select 50 x 12 year olds and 50 x 18 years olds. To make the sampling representative it would be best to use a quota system within each age group recruit children from different socioeconomic groups and equal numbers of boys and girls. Include lie scale questions to check on the honesty of participants (e.g. Do you think you behave well all the time? Anyone who answers yes is lying!). Procedure: Select participants and brief them about what will be required, including informed consent. Ask them to complete the questionnaire in a quiet classroom, on their own. Instructions on the questionnaire say This is an important questionnaire, aiming to help us to understand family relationships. Take your time and try to answer all questions truthfully and fully. The information will be used to help families understand their children better. Method of investigation: Naturalistic observation good to observe people in their usual environment. Controlled observations: Using a behaviour checklist to enable unbiased recording of behaviour. Materials/apparatus: Behaviour checklist (would contain categories such as shouts at players on pitch, shouts at individuals nearby, shakes fist/arm at individuals nearby), video recordings of a variety of fans at football matches. Procedure: Recordings made at different venues involving different football team supporters (to ensure representativeness). Observers trained in use of the behaviour checklist and then three observers record the behaviour of individual people at football matches (use three observers to ensure validity by averaging across observers). Record information relating to 50 different individuals to get representativeness. For each person observe their behaviour throughout the match. Hypothesis: TV programmes shown between 8 and 9pm contain less sex and violence than programmes shown between 9 and 10pm. Categories to use: kissing, talking about sex, hitting someone, using a gun. Sampling method: Select 20 programmes between 8&9pm and 20 programmes between 9&10pm. Observe recordings using time sampling once every minute observers tick behaviours being displayed. Procedures: Use three observers (to enhance reliability) and record sample of programmes. Observations between the three observers should be compared to see if there is close agreement.

2.

3.

4.

13

5. I would arrange to meet with the patient and administer a group of memory tests to see what capabilities he had. I would also use psychological tests to assess his intelligence and personality, to get an all round picture. I would continue to test him over a period of months and even years to check on any changes in his abilities. I would interview friends and relatives to find out what he was like before the accident and the reallife effects of his memory changes. Informed consent would be difficult if he had no memory so I would seek the consent of his nearest relative. This would involve explaining the value of the research, the amount of testing and frequency, the potential negative effects (e.g. distress from being tested), confidentiality and anonymity. Validity The tests used can be checked using concurrent validity (comparing the results of two similar tests), observations of everyday behaviour can be used to confirm (validate) the findings from tests. Interviews should enhance validity by avoiding leading questions and questions that may arouse social desirability bias.

QUESTIONS 16.12
1. (a) Most suitable measure of central tendency = mean because it takes all the values into account in the final analysis so represents all the data. Most suitable measure of dispersion = range or standard deviation because both give useful information defining the data set. (b) Most suitable measure of central tendency = median because there are extreme values which would unduly affect the mean. Most suitable measure of dispersion = standard deviation because the range would exaggerate the spread of the data, but the standard deviation would be less affected. (c) Most suitable measure of central tendency = mode because data is nominal. Dispersion measure not possible. 2. (a) Alternative hypothesis = Reaction times before drinking coffee are slower than after drinking coffee. Conduct study by using opportunity sample, test all participants reaction time (using a computer application), giving them a drink of coffee and testing reaction time afterwards. Hypothetical data: Participant 1 2 3 4 Reaction time before 10 secs 15 10 8 Reaction time after 9 15 8 10

14

5 6 7 8 9 10

12 19 15 14 16 10

9 16 15 12 17 10

Appropriate inferential test = Wilcoxon because we are looking for a test of difference for repeated measures. The data are interval so this test is acceptable. (b) Alternative hypothesis = Old people watch more hours of violence on TV than younger people. Conduct study by using an opportunity sample in several shopping centres to get a wide range of ages. Questionnaire to ask people about their TV watching habits and to collect data about how much violence they watched on average per week Hypothetical data: Old people (over 50) 10 minutes 15 10 8 12 19 15 14 16 10 Young people (under 30) 9 15 8 10 9 16 15 12 17 10

Appropriate inferential test = Mann-Whitney because we are looking for a test of difference for independent groups. The data are interval so this test is acceptable. (c) Alternative hypothesis = There is a negative correlation between reaction time and age. Conduct study using opportunity sampling in various shopping centres, by (a) record each participants age (as this is an opportunity sample may not get a good spread of ages, if you start being selective to get a good spread then it becomes more like a quota sample) and (b) measure reaction time using e.g. a reaction time test on the internet. Control extraneous variables before testing reaction time ensure that participants havent recently had coffee or any other stimulant, ensure that participants are all tested at the same time of day (time of day may affect reaction time).
15

Hypothetical data: Age 23 24 28 30 30 35 36 43 50 58 Reaction time 10 secs 11 15 8 22 15 10 14 16 18

Appropriate inferential test = Spearmans rho because require a test of correlation, data are interval so this test is acceptable. (d) Alternative hypothesis = Doctors have higher levels of stress than nurses. Conduct study by measuring stress levels in doctors and nurses in several different hospitals and doctors surgeries (use volunteer sampling). Stress can be measured by giving participants a questionnaire which asks questions about the way they respond to certain common situations, such as a patient who gets angry. Hypothetical data: showing stress scores Nurses 10 15 10 8 12 19 15 14 16 10 Doctors 9 15 8 10 9 16 15 12 17 10

16

Appropriate inferential test = Mann-Whitney test, because need a test of difference and independent groups. Data are interval so this test is acceptable. (e) Alternative hypothesis = There is a difference in the performance of the group using revision technique 1 and the group using revision technique 2. Conduct study by recruiting volunteers in a college (place ad on noticeboard), test for memory and match participants allocating one member of each pair to group 1. Hypothetical data: Matched pair 1 2 3 4 5 6 7 8 9 10 Group 1, score after using revision technique 1 10 15 10 8 12 19 15 14 16 10 Group 2, score after using revision technique 2 9 15 8 10 9 16 15 12 17 10

Appropriate inferential test = Wilcoxon test, because test of difference required and participants are related (by being matched) and therefore repeated measures design. Data are ordinal or better (interval). (f) Alternative hypothesis = People who have a pet are happier than those who dont. Conduct study by using an opportunity sample and giving participants a short questionnaire that asks if they have a pet and also asks questions to rate their happiness (such as On a scale of 1 to 10 rate how happy you feel today, where 10 is very happy and 0 is very unhappy). Hypothetical data: happiness scores Owns a pet 10 15 10 8 No pet 9 15 8 10

17

12 19 15 14 16 10

9 16 15 12 17 10

Appropriate inferential test = Mann-Whitney, because need a test of difference and independent groups. Data are interval so this test is acceptable. 3. (a) Students are more obedient with male teachers than with female teachers. (b) The male teachers might all be younger therefore greater obedience might be due to age rather than gender. The students might be mainly boys and therefore might obey men more. (c) Respond to direct request, listening to teacher, follow instructions to class in general. (NB these all permit measurement of the dependent variable) (d) Choose event sampling because the above behaviours wouldnt happen that often therefore would miss them with time sampling. Sampling also refers to selection of students could do this randomly by putting all students name in a hat. (e) Descriptive count up behaviours in each category and work out mean for male and female teachers, illustrate in bar chart. Inferential Wilcoxon test to assess the difference between males and females: selected because using a repeated measures design, data will be frequency counts of behaviours in each category (i.e. ordinal or better). (f) The conclusion is that as the result is significant this means we can reject the null hypothesis and accept that students are more obedient with male teachers than with female teachers. (g) Abstract: would contain brief details of aim/hypothesis (considering whether students are more obedient to male or female teachers, outline of procedures (observation of behaviours in classroom), results (significant difference, students more obedient to male teachers), implications (e.g. explains why female teachers have more discipline problems). Introduction: Consider research on obedience and then possible gender differences, leading up to aims/hypothesis. Method: Describe participants and target population and sampling method. Describe procedures giving sufficient detail for replication. Results: Present descriptive statistics (e.g. mean figures for behavioural categories, bar chart showing differences between male and female teachers), inferential statistics (Chi-squared test to show differences between male/female teachers and different behavioural categories). State whether hypotheses can be rejected. Discussion: State conclusions, relate findings to other research, consider methodological problems (such as observation methods), consider implications (e.g. if female teachers have issues with obedience how they might develop these). 4.
18

a. Open: What are your reasons for smoking? Closed: How many cigarettes do you smoke a day on average? b. Open: Group answers into categories and count how many in each category, or just give some of the most common answers. Closed: Can show in a bar chart, or give the mean answer. c. Advantage: People may be more willing to answer a questionnaire truthfully. Also would be quicker to give out to hundreds of participants. Disadvantage: Participants may be less willing to elaborate their answers to open questions because they have to write it out. d. To make sure that respondents were not affected by investigator bias the investigator might give them different instructions and this would lead to, for example, being more or less honest. e. One thing you might do is compare answers from smokers and nonsmokers. For example the closed question above you could put the frequency data in a contingency table and work out a 5 x 2 Chi-square. f. The questions may well reveal the aims of the questionnaire to respondents and this would affect the answers they give e.g. smokers might give politically correct answers rather than what they actually think/feel. 6. a. IV = mixed versus single-sex wards b. DV = happiness (operationalised by asking participants to rate their happiness) and health (operationalised by recording blood pressure). c. Patients in mixed wards have different happiness scores than patients on single sex wards. d. (i) Independent groups, (ii) the patients in the two different wards might differ in important ways e.g. it could be that patients who are more easy going and happy prefer mixed wards and thats why there is a difference, (iii) the researcher could randomly assign patients to ward type. e. (i) The design is independent groups and the question is whether there is a difference between the happiness scores; the data (rating) are ordinal, (ii) For 12 + 12 participants the critical value is 37, the observed value of 29 is less than this therefore significance is shown supporting the hypothesis. f. Use the mean to give an average for each ward. The mean is good because all the values of the data are used in the final calculation. Use standard deviation to represent the spread of the data for each group, this is better than the range because all the values of the data are used. Use a bar chart to represent the means for each group. g. Design: Construct a questionnaire containing questions such as: Which kind of ward would you prefer? Sam-sex or mixed sex Explain why you are in hospital. How long is your stay likely to be? 1 night 2 nights 3 nights more than 4 nights Questionnaire: Total of 10 questions so people dont find the questionnaire too onerous. Sample: Patients from several hospitals should be contacted from different areas in the UK. Used a stratified sample so that 30 patients of each sex are contacted and this is subdivided into age groups (10 aged 10-30, 10 aged 30-50 and 10 aged 50+). Ethical issues: Anonymity should be assured by only requiring sex to be recorded on the questionnaire. Procedure: Standardised instructions will contain an explanation of the purpose of the questionnaire (to assess attitudes about mixed sex wards), advise participants about how long the questionnaire will take and advise on anonymity (no names recorded) and confidentiality (data stored in protected manner). Patients will be asked to complete the questionnaire immediately.
19

No.16.12

QUESTIONS 16.13
1 One advantage of qualitative analysis is that it gains access to thoughts and feelings that may not be assessed using quantitative methods with closed questions. One disadvantage of qualitative analysis is that it is more difficult to detect patterns and draw conclusions than it is with quantitative methods. e.g. Milgrams obedience study: quantitative data = number of participants who gave full electric shock (65%), qualitative data = participants comments on why they obeyed. Banduras bobo doll study: quantitative data = aggression score, qualitative data = words they used when playing with the doll. 3 4 Opinions might be represented in a qualitative analysis by selecting common themes expressed by different participants and presenting the typical themes with examples. Triangulation refers to the practice of comparing the results from a range of different studies looking at the same topic, each study using different research methods. It is a useful way of confirming the validity of the results especially by comparing less generalisable lab experiments with less controlled qualitative research. Validity is concerned with getting at what is real, assuming that there is an objective reality out there. Qualitative researchers are more concerned with the reality as experienced by each individual and aim to be subjective rather than objective. It is a content analysis because they are examining the content of what people said, an indirect observation. It is qualitative because the data collected is not numerical. In each behavioural category instances are described rather than being counted. 7 Qualitative research may reveal some aspects of human behaviour not found through quantitative research e.g. that a person does not always disclose information about themselves just if their friend does. The first task would be to select a patient for study, which might be done opportunistically. The researcher might be interested in themes such as ability to cope, experience of being in hospital from the childs perspective, separation between child and family. The researcher would collect information from: The child his/her feelings about treatment and feelings about his/her family (how much he/she missed their family). The researcher might interview the child on a daily basis and/or could ask the child to keep a diary by talking into a recorder. Personality tests might be used to assess the child and explain his/her abiity to cope (or not). The parents - their experiences about visiting their child, and their interactions with the medical staff. Parents might be given a written questionnaire.

20

The hospital staff interviewed abut the individual characteristics of the target child and his/her ability to cope, attitudes to the childs family. At the end of the study information related to each theme would be put together and some key examples/quotes used to represent the findings.

21

Вам также может понравиться