# Question One Male participants aged 7-11 years of age will play on the road more frequently than female participants aged 7-11 years of age. B

 Page 1/3 Date conversion 07.09.2017 Size 0,56 Mb.
1   2   3

Question One

1. Male participants aged 7-11 years of age will play on the road more frequently than female participants aged 7-11 years of age.

B

1. Yes, it is significant, at all probability levels, e.g. 1%, 5% and 10%

2. Two factors are: That the data was nominal (e.g. the data was non mathematical and could not be ordered) and a test of difference was needed).

C

1. On Methodological problem is that the observer’s observations will be subjective/ biased. She may be incorrect in judging age or gender.

2. If another observer had been involved than the first observer’s observations would have been more reliable and valid. For instance two people judging the age of a child. They could then see if their scores had a positive correlation. If there was no correlation or a negative correlation than it would have been apparent that the observation techniques needed to be operationalised better to ensure improved reliability.

3. Inter-rater reliability is when two or more observers observe the same behaviour and score it. They then correlate their scores to see if they are similar, e.g. if they have a positive correlation. If they do, they can then be added together to form a mean result for that observation. If there is no correlation or a negative correlation than it is apparent the observation techniques needs to be operationalised better to ensure improved reliability

D

1. One problem with observation is demand characteristics, especially in a research situation similar to this. The children may have wondered why an adult was standing around staring at them. I would imagine being covert (hidden) would not always be possible in such an observation, where would you conceal yourself in a neighbourhood (behind a tree in a car – all a bit suspicious? This brings a second issue up, which is ethics, maybe children will be alarmed if they realise they are being watched and may not want to play outside (socially sensitive research). There are also issues of research without informed consent. Lastly, naturalistic observations similar to this type can’t be replicated.

2. The children are playing in their natural environment. Their behaviour has not been manipulated by the researcher. This makes the findings more externally valid as it representative of how real children/ girls/boys behave/play and observe road safety.

3. This research is considered scientific as the data is empirical.

• Although psychological observations try to follow the empirical method, e.g. having a hypothesis that can be falsified, operationalising variables and training observers in their techniques, it could be argued that there is too much that cannot be controlled: extraneous variables, observer effects/bias, demand characteristics, procedure cannot be replicated, etc. Therefore we cannot really be sure that the result is worthy any kind of scientific label.

• Others researchers have criticised this view of science when it concerns people/psychology. The mains problems are that: Objectivity is impossible. It is a huge problem in psychology, as it involves humans studying humans, and it is very difficult to study the behaviour of people in an unbiased fashion. Moreover, in terms of a general philosophy of science, we find it hard to be objective because we are influenced by a theoretical standpoint (Freud is a good example of this). Also, the observer and observed are members of the same species are this creates problems of reflectivity.Human beings are unlike physical objects so total control of variables would be impossible (you would have to clone people and bring them up identically).

• We need to keep trying to develop scientific methods that are suitable to studying human behaviour - it may be that the methods adopted by the natural sciences are not appropriate for us.

• The data is nominal. Nominal data is simple data. Usually it is subjective categories/observations that have no real scientific or mathematical meaning. For instance in the above research example we wouldn’t know much from the data. How frequently each child played in the road for how long and which age group were the most persistent. Psychologists try and quantify their observations turn them into mathematical data. The difference between males and females or playing in the road or not is not mathematical so how can they be measured mathematically?

E)

Please note this can’t be a correlation.

Design: Non experimental questionnaire survey. Variables: Gender and scores on road safety questionnaire. Questionnaire will be quantitative data measuring attitudes towards road safety.

Controls: make attitude scale an even number so pps can’t choose the middle value. Standardised instructions, same time to complete questionnaire.

Participants: Sixty children aged 7-11, half were female and half were male, mixed ethnicity, all middle class from an Orpington local primary school. Obtained via volunteer sampling, e.g. letter sent to parents who volunteered their children.

Material/apparatus: Questionnaire, pens, standardised instructions, mark scheme, consent form for parents, debriefing.

Procedure and ethics: Fully informed consent. Parent’s sign consent as kids under 16. Children come in one at a time. Read standardised instructions. Ask if they are understood. Complete questionnaire. Debrief parents and participant.

Question 2

A

i) The results in table one suggest that the depressed group took slightly longer than the non depressed group in estimating how long it took them to complete a task. This difference was very small, 16.1 seconds. As the method of central tendency used was the mean we can not be sure that it was not affected by anomalies/outliers.

B

i) Not statistically significant at the 5% level means that there was a more than 5% chance that the results did occur because of the experimental manipulation of the IV. In other words, the experiment did not work at the significance level chosen, the null hypothesis cannot be rejected and we can presume over 95% of the participants did not perform in the experiment how it was hypothesised/predicted they would.

ii) Choosing a 10% significance level increases the chance of making a type 1 error, e.g. rejecting your null when you should have accepted your null. This is because there is a ten in a hundred risk that any results at this significance level are will be due to chance. Sometimes psychologists go to the other extreme and choose a 1% significance level, this means reduced type one errors but increased type two errors, e.g. accepting the null when it should have been rejected.1% levels are too stringent and 10% are not stringent enough. Choosing a five percent level decreases the amount of type 1 and type 2 errors.

C

ii) Needed a test of difference, the data was interval/ration e.g. could be ordered and gaps were mathematical and the groups were independent.

D)

The researcher decided to use a neutral task, e.g. four letter strings that did not combine to make words so that participants could not associate the meaning of a particular word with an event or memory which could then alter their behaviour and become a confounding variable. Words like ‘Dark’ may conjure up depressive thoughts and may make participants evaluate time going more slowly. On the other hand words like ‘walk’ may make some participants feel more energetic and happy and evaluate time going more quickly.

E)

Informed consent; should you tell participants the hypothesis? Should you tell participants that they will take a questionnaire measuring their depression? Both could cause demand characteristics and thus affect the internal validity of the study.

Alternatives would be to use prior general agreement or presumptive consent. This still leaves the problem of psychological harm as some participants may be disturbed to learn they are depressed.

F) i and ii

The researcher divided the participants into depressed or not depressed groups by using the median. Those above the median were put into the depressed group.

This not only an unfair/unethical method of assessing depression but it is also invalid. Medians just use the middle number of scores.

We know from the information available, that participants were taking a standardised depression scale where a high score indicated Depression.

Let’s imagine the depression scores were out of 120 and that people were rated depressed if their score was 70 and above.

Let’s suppose ‘our’ participants got, not only a very similar set of scores but also a very middling set of scores that were neither *high/low and were within the normal range, e.g. under 70. For example: 45, 48, 50, 55, 56, 58, 58, 59, 59, 99.

*This might actually be a possibility as the sample were motivated enough to apply and participate in study; unlikely to be depressed then?

• This would make the median 58

• Someone would be deemed depressed if they scored 58 and above according to the above median and not 70 according to the scale’s instructions.

• This means half the participants have been inaccurately judged as depressed because they should have got a score of 70 and above to be deemed depressed.

• These participants will be given inaccurate information about their mental health in a debriefing session (unethical).

• The study will not work properly because it is not testing what it is supposed to, e.g. depressed participants measuring time against non depressed participants. It is therefore internally invalid.

• A better method for allocating participants to groups would be to put those with scores above 70 into the depressed group and below into the non depressed group. Even if group numbers were uneven this would not affect statistical analysis, e.g. tests can compensate when groups are unequal in size.

ii) sample and generalisability

G)

The case study is an idiographic and qualitative approach to research so it is high on internal validity but low in external validity; you can not apply or compare one person’s experience to anyone else’s. However, learning rich detailed information about an individual’s experience of time evaluation and mood can help a psychologist understand and target specific problems for a specific person. Some psychologists would argue that we need to be researched like this anyway because we are individuals.A person’s subjective experience of the world is an important and influential factor on their behaviour. Only by seeing the world from the individual’s point of view can we really understand why they act the way they do.’

Some psychologists would use information from a variety of different case studies because it can improve their knowledge about their research interest in a variety of ways:

• How to design a study/experiment,

• Alert them ideas they hadn’t considered or anticipated as being important.

This is amore inductive approach to science, e.g. making predictions based on observations

H)

i) Title, Abstract, Introduction (Research Justification and History, Aims, Hypothesises), Method (Design, Participants, Apparatus/Materials, and Procedures): Results (Descriptive Statistics, Inferential Statistics) Discussion, Refs, Appendices. The answers in red are what you need for full marks. The black stuff is icing on the cake!

ii) Peer review is when:

• Psychologists send their unpublished research to other psychologists for review and judgements.

• This is done by psychologists in the same field as the researcher so that more reliable reviews can be made.

• Advantages: It means qualified psychologists can weed out research that is unethical, invalid or unreliable. Poor research essentially.

• Disadvantages: The research may be so unique there are no experts; Researchers may not just be peers they may be adversaries in the exact same research area. One researcher may criticise research that invalidates or disputes theirs because of jealousy loss of credibility etc.

Question Three

A)

Advantages: Volunteers samples are cheap, quick and easy to obtain.

Disadvantages: Not representative of the population. They may have many of the same traits/personalities, e.g. time to volunteer (so maybe Mothers, old people unemployed, or perhaps people who volunteer have helpful inquisitive natures.

B)

The graph suggests that participants in group 1 (therapy A) scored higher on an anxiety scale assessment before and after their therapy than group 2 (therapy B). It also showed that participants in group 2 (therapy B) seemed to have a more marked improvement in their anxiety scale assessment than group 1 (therapy A).

C)

Man Whitney U because a test of difference was needed, the groups were independent, the level of measurement was ordinal (gaps between variables could be ordered but were non mathematical).

D)

i) Groups that are matched on variables that are considered important for the research being undertaken so they do not confound the result, for instance in the above research scenario: matching participants on age, past experience of therapy and/or flying, personal history of phobias in the family.

ii) Matching key variables would have meant that individual differences such as past experience of therapy or flying would have not confounded the result.

E) Actually going through the flight process. As the pp’s were not matched perhaps some of the pp’s fear was socially learned from parents and they had never been on planes and other pp’s phobia was from actual flying experience, e.g. travelling through turbulence. Thus the pp’s with no flying experience actually realised their cognitions had been different to the reality of flying and felt better solely for this reason. By using a matched pair design or only recruiting pp’s who had never flown or with similar flying experiences.

F)

Putting self confessed flying phobics through a flight simulation. Even though they agreed (fully informed consent) it still may have caused psychological harm (anxiety). Also deception, telling pp’s the plane would actually take off when it was never going to.

G)

The experiment was conducted with a real life scenario actually going to an airport and going through the procedures of take off so it had Mundane realism (e.g. can the results of a study be applied to everyday life? E.g. How often would we be asked to give electric shocks as a punishment for giving wrong answers in real-life -as the participants did in Milgram's (1963) obedience to authority study)?

It also had ecological validity (can the results of a study be applied to different environments?). Yes! We could say that most aeroplanes take off this way and pre fight conditions in most airports are similar so the study is ecologically valid.

Therefore the experiment is externally valid.

NB. It's possible for a study to lack both ecological validity and mundane realism or lack one but not the other.

H) validity

1) Ask the participants if they believed they were really going to take off. The problem with this is they may lie either to please the researcher or to preserve their status.

2) Observation/video of pp’s for signs of stress as an indication of belief that the plane would take off. Problem, not all people visibly show stress/anxiety.

Reliability

1) Replicate study to see if results are similar (test retest). Problem, you may get similar proportion of people displaying demand characteristics.

I)

Design:

Repeated measures experiment. IV Therapy A and Therapy B. DV scores on an anxiety scale.

Controls: Counterbalancing to avoid order effects of therapy B or A. Standardised instructions, all participants have similar flight and therapy histories.

Participants: Sixty PPs ½ male and ½ female. All were aged between 25 -35. All were British. Recruited through advert in Daily Mail (conservative paper so perhaps can infer pp’s were middle class and conservative).

Apparatus/Materials: Not asked for in the question so I’ll leave it out.

Procedure and ethics: Prior general agreement, (as deception of flight not taking off), standardised instructions, give all pp’s anxiety assessment, randomly allocate for counterbalancing group, flight simulation then therapy A or B depending on counterbalanced group followed by anxiety assessment. Then therapy A or B depending on counterbalanced group followed by anxiety assessment. Debrief.

Note If it is a repeated measures design you should mention the results could be confounded by ‘order effects’, e.g. the order the participants take the conditions. There are two possible outcomes here

A) Participants could either be unsure of what they have to do in the first condition and do badly because they don’t know what they are doing then in the second condition they become experts because they now know what to do. The results then could show condition two doing better. You (the researcher) think it’s to do with the IV but it is to do with order effects.

B) Or, participants could be eager in the first condition and then become bored in condition two (they have to do the same thing again!). The results then could show condition one doing better. You (the researcher) think it’s to do with the IV but it is to do with order effects.

C) Solution: Counter balance! Make ½ the participants do condition one 1st and ½ the participants do condition two 1st then. Then ½ will do condition two second and ½ will do condition one second. The results are counterbalanced.
 Participants IQ result in Music with words condition 1 IQ result in Music without words condition 2 1 99 34 2 88 23 3 90 54 4 100 34 5 78 34 6 98 56 7 89 21 8 98 13 9 87 65 10 76 21 NOT BORED BORED but because the participants are eager to do well This condition does worse because the participants can't be This condition does better not because of the IV bothered to do an IQ test again. The researcher thinks it but because the participants are eager to do well because of the IV. Counterbalanced The group is split into 2. ½ the group do condition A 1st & ½ the group do condition B 1st Participants IQ result in Music with words condition A IQ result in Music & no words condition B 1 99 34 2 88 23 3 90 54 4 100 34 5 78 34 IQ result in Music & no words condition A IQ result in Music with words condition 6 98 56 7 89 21 8 99 13 9 87 65 10 76 21 NOT BORED BORED

Question Four

A)

Questionnaires

Advantages: Cheap, quick and easy to conduct and create. It is usually quick to train people to administer. If quantitative, questionnaires can be easy to score. Good for preliminary research on an area.

Disadvantages: Depends on the method of delivery but if asked directly and filled in by an interviewer or the person has to give their name, there may be demand characteristics. If the questionnaire needs to be returned there is a low response rate. People can misinterpret questions. Quantitative questions lack validity. There can be biased samples as only some people will answer questionnaires.

B)

i) A pilot study is a preliminary study or trial often carried out to predict snags and find out about the main features of a main research study that will follow it. A trial run basically. Piloting is trying out the scale or research design on a small sample.

ii) So that the researcher could see if participants understood the questions, e.g. they were not ambiguous (e.g. ‘do you smoke: rarely, often, sometimes is ambiguous as different people have different ideas about these words – rarely to me is once a year) Also to check that questions were not leading, confusing, unethical and full of jargon. Asking pp’s under 18 about smoking is socially sensitive research.

To see if the form design was appropriate, did participants need more space etc?

C)

i) Questions: 1, 2 and 3 are quantitative maybe 1st part of question 5.

ii) Questions: 4, 5 and 6 are qualitative. Five is a double barrelled question. The 1st part could be quantitative and the 2nd part is qualitative.

iii) Because it is an open ended question. The participants must come up with an answer themselves, rather than tick from a selection of pre set response choices that can be turned into numbers and quantified.

D)

i) So many of the questions are badly designed or written. I have highlighted one already in B ii (rarely, sometimes, and often)

Also, Q5 is double barrelled, two questions in one sentence.

Q6 is a leading question. It is making the participant write a paragraph stating that smoking is bad for your health. I would rewrite this as two questions:

1) Do you think smoking is dangerous to health? YES NO UNSURE

E)

Making sure that all pp’s did not have to give personal information such as names, addresses etc as smoking is anti social and unacceptable people may want to keep their smoking habits confidential, it is socially sensitive research. Also pp’s may be under 16 and smoking is illegal so confidentiality is essential. I would ask for fully informed consent to ensure that people were fully aware of the nature of the questionnaire so that they did not part in something that may compromise their safety or admitting to breaking the law.

F)

Code your data. In qualitative studies, coding means identifying themes within your questionnaire that relate to the research questions in your study. Themes are common ideas and patterns that you come across repeatedly as you read the data you've collected. i.e., identify certain ideas and themes in answers that involve qualitative data. For example peer pressure would arise as a category in Q4. You will likely have to read through your data multiple times to identify all of the themes. Interpret your data by attaching significance to the themes and patterns you've observed. Write lists of key themes and review the data again. Consider alternative explanations by looking for differences in responses or observations that you recorded in your data collection.

Qualitative data are limited in their external validity or their ability to be generalized to the general population. A quantitative study may be necessary in order to have good external validity.

Qualitative analysis often requires more time and effort than quantitative research. The fact that data collection, analysis, and reporting findings overlap complicates matters further. Some specialized computer software programs exist for qualitative data. However, these programs help organize the data rather than conduct analyses.

Analyzing qualitative data is a subjective process, making researcher bias a potential concern. Be aware of your own biases throughout the analytical process.

G)

The main method used are interviewing people about their drug/drink habits and using correlational studies. The method is non experimental so cause and effect can not be demonstrated. Some researchers do not see a problem with this kind of research. The laws of cause and effect work for inanimate objects, if one billiard ball collides with another, you can predict with a fair amount of accuracy the final resting place of each. Moreover after the initial collision they no longer influence each other. Living systems are another matter. If I kick a dog, I could calculate how far the dog should travel in a particular direction, given its size and weight. The reality would be a bit different though- if I were foolish enough to kick a dog, it might turn around and bite my leg; the dog’s final resting place is very unlikely to have anything to do with Newton’ Laws of motion. Human beings are complex – many things happen simultaneously. You cannot predict exactly will occur, because one person’s response influences the other person’s communication. The relationship is a loop; we are continually responding to feedback in order to know what to next. Focusing on one side of the loop is like trying to understand tennis by studying only one end of the court.

This leaves you with the question of whether you think non experimental research is worth doing as it is non-scientific. For this reason psychology has a problem with always following a truly scientific procedure when they are not studying a biological aspect of behaviour. Whether this means that YOU think that research that does not show cause and effect is rubbish or that it is still worthwhile……for instance, if you build up enough non experimental methods and they indicate links to certain causes (Triangulation). It is up to YOU to dispute and make commentary on whether you think it is worthwhile.

Alternative Hypothesis: There will be a positive correlation between the age of participants and the number of cigarettes they smoke; the older the participants are, the more they will smoke.

Null Hypothesis: There will be no correlation between the age of participants and the number of cigarettes they smoke.

One tailed hypothesis because previous research has indicated people will become more addicted to nicotine the older they get.

Design: Non experimental, correlation. Co Variables: Age and no: of cigarettes smoked, operationalised as a questionnaire asking pp’s about age and smoking habits. Data will be quantitative and interval. Controls: Standardised instructions, Slot box so anonymity can be maintained and thus avoid demand characteristics, same for no name on form.

Inferential test: Spearman’s row as data will be interval (age and no; of cigarettes is mathematical data that can be ordered). I need a test of correlation.

Procedure: pp’s in one at a time. Fully informed consent form and sign, Read Standardised instructions, give examples of the questions and where you are and how long it would take, debrief and thank.

Hypothetico-deduction: is the view that science proceeds by deriving hypotheses from theories, which are then tested for truth or falsity by observation and experimentation. It is the opposite of induction, which proposes that theories can be derived from observations.

Question 5

A) Identify the type of experimental design used in this study. (1 mark)

This is a question about the design of the experiment, e.g. whether it is: Independent group design, Repeated measures or matched pairs. You will not get marks for saying it is a laboratory experiment or field btw.

The answer is: Independent group. Individual differences (contextualised)

B) Identify one extraneous variable that the investigator addressed in the procedure for the study and explain how it was addressed. (4 marks)

You only need to suggest one thing here AQA like you to talk about ‘order effects and suggest counterbalancing, if it is a ‘Repeated measures design’ and randomly allocating if it is an ‘Independent group design.

II) If it is an independent group design and you do not randomly allocate participants to conditions, the groups will be biased; everyone should have a chance of being selected.

A) Solution: Randomly allocate participants to conditions by drawing names from a hat (AQA wants you to suggest how you would randomly allocate to conditions).

III) If you can’t remember any of the above, you could mention that the sample have prior knowledge of whatever it is that is being tested or that they are too different in terms of age, IQ, or whatever variable you think is relevant or could confound the DV/variable being rested.

A) The answer I would give is: That some of the participants may have knowledge of the tourist industry and or attractions in their better town better then others.

B) I would suggest controlling (making sure participants had the same level of key variable) for knowledge of the town and or tourism when selecting participants.

C) Name an appropriate test of statistical significance for analysing this data. Explain why this would be a suitable test to use. (4 marks)

• You would need to know the following to answer this question:

• What the DV is or how you are measuring the results (if non experimental).

• What level of precision/data you have: Nominal, Ordinal, Interval/Ratio etc

• Your method. Whether you are doing an: experiment, correlation, observation etc so that you know whether you need a test of: Difference or correlation.

• Your experimental design: If it is repeated measure/matched pairs or Independent group design.

Is your investigation a study of?

I chose a Man Whitney U test because my experimental design was an independent group design, my data was interval/ratio (e.g. it had mathematical gaps between variables), and I needed a test of difference.

D) What does p ≤ 0.05 mean?

• p stands for probability

• means less than

• when it is underlined, it means Less than or equal to ... ≤

• >means greater than

• when it is underlined, it means more than or equal to ...

• 0.01means 1%

• 0.05 means 5%

• 0.10 means 10%

Answer: p ≤ 0.05 means the probability of the result occurring by chance is less or equal to 5 percent. That there is a ninety-five percent chance the result was not due to chance and is therefore statistically significant. Or you could just say, ‘The probability of the results occurring by chance is equal to or less than 5 times in 100.’

For the second 2 marks describe the increase in probability would increase type 1 error and vice versa and why.

E) population validity – ability to apply findings to different people, ecological validity – ability to apply findings to other settings (contextualized)

F) With reference to the data in Table 1 outline and discuss the findings of this investigation. (10 marks)

AO2 = 4 marks Outline of findings of the investigation

AO3 = 6 marks Analysis, evaluation and interpretation of other's methodology and the impact of findings

Answers should describe the overall results and make reference to:

• The average scores for 2 conditions.

• Range or standard deviation (SD) for each condition

• Explain what the mean and SD seem to tell us then basically you should A02 the research just like you would if you were doing it for an essay. Consider anything that may make you doubt the findings:

• Internal validity (e.g. demand characteristics, experimenter bias, bad tests or methods of testing the DV, poor or no operationalisation of variables and/or IV and DV, realism of task WAS IT ATIFICIAL OR TRUE TO LIFE).

• External validity (e.g. ecological validity, population validity (biased sample)

Answer: (There are lots of variations on the following way you can discuss this 1st bit.

The mean in the ’working alone condition’ was 14 compared to 8 in the ‘working in a group’ condition. This is almost twice the amount. It does not tell us about extreme outliers (values) though.

The SD is a stronger descriptive statistic as it tells how much the participants deviated/ ranged from the mean. In the ’working alone condition’ the SD was 1.09 smaller then the working in a group’ condition. The smaller the SD, the more similar the sample’s result is to each other. This instills confidence that the IV may have affected the DV. Moreover, if we assume there was a normal distribution, which there would be as we were given the average (the average is when the mean mode and median are the same and means that there is a normal distribution), the SD in the ’working alone condition’ was small, it showed that 68.75% of the sample scored between 15.89 words generated and 12.11 words generated compared to 68.75% of the sample in the ‘working in a group condition’ scoring between 10.98 and 5.02. However, both groups should have similar SDs, if one is very small and the other larger then we would assume that one group had more reliable findings (e.g. the working alone group). As the ‘working in a group condition’ had a bigger SD we could assume there was less reliability because there was more variance in their scores.

Inferential stats showed this result was significant.

Realism of the task was good as many people may have experienced ‘brainstorming’ activities in school or the work place (GOOD mundane realism and ecological validity).

However, there was low reliability of identifying who generated each an idea when participants were working in groups. Do you think the researchers’ use of filming participants was valid and/or reliable?

What constitutes "an idea". There was no operationalisation!

G) The psychologist noted that younger participants seemed to generate more ideas than older participants. Design a study to investigate the relationship between age and ability to generate ideas. You should include sufficient details to permit replication, for example a hypothesis, variables, detail of design and procedure, sampling. (12 marks)

This can not be an experiment! As there is only one condition both groups are doing the same thing!! Also if you have two distinct age groups you can not randomly allocate to conditions.

Design

Non experiment, I would do a Correlational study. The co variables would be age (Participants who are aged between 20-30 years and participants who are aged between 40- 50 years old) and number of ideas generated (this will need operationalizing e.g. getting children to eat healthy meals)

The researchers will not count an ‘idea’ on the basis of its quality/credibility, as that is subjective and could be different in quality due to individual differences such as intelligence etc. The researcher will merely count the ideas each person comes up with. The data is therefore interval ratio. A spearman’s rho would be used for inferential stats.

Controls: Standardised Instructions, no names on forms to avoid DC. Slot boxes to avoid DC.

Hypothesis

‘There will be a negative correlation between Participants who are aged between 20-30 years and participants who are aged between 40- 50 years old in the number of ideas they can generate on how to get children to eat healthy meals.’

This hypothesis is one tailed. Significance will be at 5%

1   2   3