# Diagnosis, Therapy, Test – Marking Feedback: Research Methods

A Questions: AimsB Questions: VariablesC Questions: HypothesesD Questions: ExperimentsE Questions: Experimental DesignF Questions: Reliability & ValidityG Questions: Extraneous VariablesH Questions: EthicsI Questions: Descriptive StatsJ Questions: GraphsK Questions: DistributionsL Questions: SamplingM Questions: CorrelationsN Questions: ObservationsO Questions: Self ReportP Questions: Choosing StatsQ Questions: Sign TestR Questions: SignificanceS Questions: Case StudiesT Questions: Content AnalysisU Questions: Report WritingV Questions: Design a StudyW Questions: Peer ReviewX Questions: Economy

## Research Methods Booklet

### B Questions: Variables

RMAa

Diagnosis = Dependent Variables

Dependent Variables are the variable we are measuring. Students will often get these wrong by:

1. Getting them mixed up with independent variables
2. Not providing a unit of measurement
3. Not thinking about the exact data collected from EACH participant. Instead putting ‘average’ when they didn’t average anything.

Therapy: Check for the above errors. Is it definitely what was measured? Did you put a unit for example seconds, score, metres, rating, percent? Did you add a word like ‘mean’ or ‘average’ that wasn’t supposed to be there. Follow an example of “what result would be written down for each participant – for what task” e.g. Number of correct words recalled – in a test of memory OR seconds taken to complete – a counter moving task.

### E Questions: Experimental Design

RMEa

Diagnosis = Counterbalancing

When they ask about counterbalancing they are really testing your understanding of order effects and want to see how well you really understand experimental design. There are a few things they are looking for:

1. Do you understand it’s for repeated measures design.
2. Do you understand that you cannot get rid of or control for order effects all you can do is balance them.
3. They want to see you understand the processes enough to explain how it achieves this

Therapy: Describe what happens in a counterbalanced design but include the order effects. For example half of participants do condition A then B meaning they get bored in condition B and the results go down. The other half…

### G Questions: Extraneous Variables

RMGa

Diagnosis = Identifying an Extraneous Variable

When they ask about extraneous variables they are looking for you to identify either a participant variable (something about the people used) or a situational variable (something about when/where it was done). This should be something (anything) which could impact the DV that IS NOT the IV. They are also testing whether you realise which of these is/isn’t relevant. In related designs the participant variables are not relevant but situation are. In unrelated designs they are very relevant. Student tend to lose marks for:

1. Being too generic and saying “participants” or “situation” it needs specific things
2. Providing an irrelevant extraneous variable by not fully reading stem. This is sometimes using the IV, or something which would not effect the DV in this case, or not looking at design.
3. Picking something it would be hard to control.
4. Picking something already dealt with in the design/stem. Normally saying order effects when its been counterbalanced.

Therapy: Check if it’s related or unrelated. Check if it’s been standardised/counterbalanced already. Then choose a relevant extraneous variable to address the question with. For IDENTIFY say what it is clearly and in specific. For EXPLAIN, say why/how that would effect the DV. For DEAL/CONTROL say exactly what you would need to do to stop it effecting the DV, explaining how it would solve the problem.

RMGb

Diagnosis = Investigator Effects

Investigator effects are any ways the researcher can unintentionally effect the outcome of the research. You will normally be asked to define and give an example from the study. Students lose marks by:

• Being too generic and not giving specific things like expectancy bias
• Not being able to explain how that might happen.
• Confusing it with participants effects.

Therapy: Check you answer is actually about the behaviour of the investigator not the participants. The easiest investigator effect is expectancy bias, due to knowing what I want/hope to find this will impact the way I behave. For EXAMPLES think about what the investigator expected to find, then consider ways they could inadvertently make that happen, either in behaviour, what they say, what the record, how they interpret things. For CONTROL think about ways you would be able to eliminate expectancy? This has to be Double Blind.

RMGc

Diagnosis = Pilot Studies

Pilot study questions are about realising why we do them and how we do them. Students usually lose marks because:

• They link it to reliability and in fact they would help with internal validity.
• They only say it “tests everything works” rather than being able to give specific examples of the kind of things which can go wrong.

Therapy: Give an introduction which states pilot studies increase internal validity by ensuring we are testing what we intend to in an accurate way. This is by making sure the study works as intended. Follow this up with specific ideas of what might need to be checked (normally the amount of things there are marks, so 4 markers is an intro plus 3 specifics). Think about time, equipment, resources, for each talk about what you need to check and what could go wrong. For example if you are presenting things for participants to read/watch pilot studies can be used to check that materials presented are clear enough to understand. If they are presented too quickly they may not be seen/read, if they are presented too slowly they may become to easy.

### H Questions: Ethics

RMHa

Diagnosis = Identifying an Ethical Issue

When they ask about ethical issues they are looking for you to identify something which could reasonably have effected the study they have given you. This should also include clear terminology. Student tend to lose marks for:

1. Not using key terminology
2. Confusing ethical issues, ethical guidelines and ways of dealing with ethical issues.
3. Picking something it would be hard to deal with (if this is required as a follow up question.
4. Being too generic and saying something like “informed consent, deal with it by getting consent” this is a guideline not an issue as such, and is a too simplistic and generic way of dealing with it, there were, after all presumably reasons why the consent was not that easy to obtain.

Therapy: Choose one/more of the following Ethical Guidelines/Issues: Informed Consent, Deception, Protection from harm, Confidentiality. For IDENTIFY state it with max 1 sentence explanation. For EXPLAIN, say why/how it was an issue in this study to do this. For DEAL/CONTROL say exactly what you would need to do including terms. For example, Informed consent due to it being a covert naturalistic observation they need to be unaware they are being studied. This can be dealt with by placing signs in the vacinity saying a study is taking place l, general prior consent and then approaching participants afterwards and informing them of what happened and asking them for their retrospective consent, giving them the right to withdraw their data.

RMHb

Diagnosis = Consent Forms

When asking about consent forms they are checking if you really understand what full and informed consent is all about. It should be written as a form for a participant to read and then sign inorder to agree to take part. Student tend to lose marks for:

1. Being too generic and saying “participants” or “situation” it needs specific things
2. Providing an irrelevant extraneous variable by not fully reading stem. This is sometimes using the IV, or something which would not effect the DV in this case, or not looking at design.
3. Picking something it would be hard to control.
4. Picking something already dealt with in the design/stem. Normally saying order effects when its been counterbalanced.

Therapy: Check if it’s related or unrelated. Check if it’s been standardised/counterbalanced already. Then choose a relevant extraneous variable to address the question with. For IDENTIFY say what it is clearly and in specific. For EXPLAIN, say why/how that would effect the DV. For DEAL/CONTROL say exactly what you would need to do to stop it effecting the DV, explaining how it would solve the problem.

### I Questions: Descriptive Statistics

RMIa

Diagnosis = Interpretation of Mean/SD

Interpret question are usually 2-3 marks per data point (average/SD). Students will often only pick up half the mark available because:

1. They do not provide the raw data.
2. They only provide the raw data
3. They do not interpret standard deviation.

Therapy: Mean is about how well each group did on average, it tells you who did better. State which group did better, how you know this (data), and what this means for the hypothesis. Standard deviation is about spread of data or variation, it tells you about how individuals differed. State which group is least consistent, how you know this (data), and what this means for the hypothesis.

### J Questions: Graphs

RMJa

Diagnosis = Graphs

Graph questions are usually between 3-4 marks and are looking for you selecting the right kind of graph, labelling it correctly and drawing it precisely. Students lose marks for:
1.Most often drawing/selecting the wrong graph.
2.Labelling the axis too vaguely for example ‘age’ or ‘condition’

Therapy: Check for the above errors. Try doing a very quick sketch of the different graphs with the data you have. Which seems an easy and sensible way of representing the data? Why? When labelling the axes it will be the OPERATIONALISED variable labels you need to use. Specific conditions, units of measurement. Normally measurement on Y conditions on X.

### M Questions: Corrleations

RMMa

Diagnosis = Correlation Hypothesis

Correlation Hypothesis are testing 3 things. Whether you realise it’s relationship rather than difference and how to make it either directional or non-directional. Also can you operationalise the co variables. Student tend to lose marks for:

1. Saying “difference” or discussing a difference when it’s relationship. This equals instant zero marks.
2. Predicting a direction when it asks for non directional or vice versa. This equals instant zero marks.
Not fully operationalising variables.

Therapy: For a non directional hypothesis you should say “There will be a RELATIONSHIP between ____________ and _____________.” Where the blanks are fully operationalised giving the exact measurements taken from each participant. For a directional hypothesis you are adding the word positive or negative before the word relationship.

RMMb

Diagnosis = Correlation Graphs / Scattergram

Graph drawing questions are about accurately drawing the graph and labelling axes. Students tend to lose marks for:

1. Not including a title.
2. Not labelling axes, or labelling them vaugly rather than precisely with units of measurement.

Therapy: Ensure you’ve labelled both axes with fully (normally 1 mark for each). Ensure you have titled the graph this should be “a graph to show…” and includes both the co variables fully operationalised.

RMMc

Diagnosis = Correlation Conclusions

Drawing conclusion questions are about stating precisely what a graph shows (where appropriate stating actual data). This is normally 1 or 2 marks. For 3-4 marks this needs to go back to the conclusions in terms of the hypothesis. Students tend to lose marks for:

1. Saying what the graph is showing (in correlation this will state direction and strength) but not linking it back to the hypothesis.
2. Saying what the graph means in terms of the hypothesis but not backing this up with evidence from the graph.

Therapy: What does the graph showing (in correlation this will state direction and strength)? What does this mean in terms of the hypotheses. This is restating as a directional hypothesis in present tense. E.g. “There is a positive relationship between X and Y. As X increases so does Y”

### V Questions: Design a Study

Diagnosis = Design a Study (Gym Observation)
These questions in the exam are worth an entire grade on their own!  They have caps in the mark scheme so for each bullet point not attempted you automatically lose 3 marks. Student lose marks in a design a study questions because:

• They do not complete one or more sections (often an automatic cap)
• They do not given sufficient procedural explanation that someone could attempt a “reasonable replication”
• They do not justify their ideas
• They given generic answers not anything specific to the study in question

Therapies:

RMVa – PSY2 A Level 2017 Gym Observation

RMVai Pick a type of observation to use, choose between each pairing:

• Naturalistic or controlled, describe what you would be able to do/why that one? Avoid just saying it will be natural/controlled but say how you would achieve it and why that matters IN THIS CASE.
• Overt or covert, describe what you would be able to do/why that one? Avoid just saying it will be ethical/avoid Hawthorne but say how you would achieve it and why that matters IN THIS CASE.
• Participant or non-participant, describe what you would be able to do/why that one?

RMVaii Pick some behaviours you are interested in. They are testing if you can go beyond grand concepts like exercising or socialising. Be specific.  It is behaviour categories so they need to be very specific behaviours, the hypothesis is about how people spend their time, 2 “exercising” behaviours and 2 “socialising” behaviours is enough.

RMVaiii Pick EITHER time or event sampling, describe what you would be able to do/why that one? Again, ensure you describe how you would do it, not simply that you would.

RMVaiv Read the question carefully here “reliability of the data collection could be assessed” this is asking for you checking that the data you have collected is consistent. You need to use inter-observer reliability but describe the whole process not just name it. What do the observers doe before, during and afterwards?

## Research Methods Specifics

RMS01 – PSY2 A Level 2017

Diagnosis = Percentage Decrease

This is a standard GCSE maths question which is assessing your numeracy skills. As with maths you need to be able to remember how it is calculated and be able to estimate if you answer is sensible.

Therapy: Step one: which of these seems sensible? 90%, 50% 10%? Should give you an idea what number you are roughly aiming for. Calculate the following formula (difference÷original)x100. Leave as 3sig.fig.