Research Methods 1 Flashcards
Independent variable
This is the variable that changes OR is manipulated (by the researcher) to see if it has an EFFECT on the DV.
Dependent variable
This is the variable that the researcher measures. To see the effect of the IV
Operationalising
Operationalising means to be specific and clear when defining the independent and dependent variable in an experiment, to make it easier to measure.
Why is operationalising important?
increases replicability/reliability
Aim
This is a general statement of what the researcher intends to investigate, the purpose of the study.
Always start your aim of with the words ‘to investigate……..’
Hypothesis
This is a specific, testable statement in which the researcher predicts what will happen between the variables.
Directional (one tailed) hypotheses:
The researcher predicts the direction that they think the results will go in.
There will be significantly higher/faster/more (operationalised DV) when participants (one condition of IV) compared to when participants (other condition of the IV)
Non-directional (two tailed) hypotheses:
The researcher does not predict the direction that they think the results will go in.
There will be a significant difference in (operationalised DV) when participants (one condition of IV) compared to when participants (other condition of the IV)
When do you use a directional hypothesis and when would you use a non-directional hypothesis?
Directional:
All past research has shown
Previous research has shown
Non-Directional:
New area of research
Limited previous research
Previous research findings have been mixed (some research has found….)
Extraneous Variables
unwanted extra variables other than the IV that could have an effect on the DV, lowering the internal validity of study
Causes of EV
- Situational factors - things to do with the environment the research is carried out in e.g. task difficulty OR standardised instructions.
( 2 tasks= same difficulty, time/ 1 task= diff ppts in each) - Participant variables - things to do with the Ppts e.g. ages, gender, IQ, eye sight, hearing, personality (introvert/extrovert). These can only be used for an independent groups design (demand characteristics)
- Experimenter variables - things to do with the researcher e.g. investigator effects.
(appearance, bias design (diff), interpret ppts behaviour)
Identify 1 EV and explain why it should be controlled (3 Marks)
- name
- explain with links
- state why should be controlled, impact on DV
Identify 1 EV and explain how it can be controlled (2 Marks)
- name
- state how you would control it - what would you change
Confounding Variables
When an EV is not controlled for and it does affect DV and so becomes confounding variable
Experimental Method - Lab
An experiment conducted in a lab which is a highly controlled environment, where the researcher manipulates IV and measures the effect on the DV
Lab AO3
:) High in reliability as the experiment (CONTEXT- into…) can be easily repeated in the same Conditions (CONTEXT), to check for consistent results
:( Lacks Ecological validity as it’s carried out in an artificial environment (CONTEXT- where) therefore it is difficult to generalise the findings (CONTEXT- of what…) beyond the setting of the study, lowering the external validity.
Experimental Method - Field
An experiment which takes place in a natural environment such as an office or school where the researcher manipulates IV and measures the effect on the DV
Field AO3
:) High ecological validity as it is based in a real life setting therefore it is easier to generalise the findings beyond the setting of the study to other similar settings (CONTEXT- example) increasing the external validity.
:( Low in reliability as the environment is natural (CONTEXT- where), it is difficult to repeat with the same conditions to check for consistent results.
Experimental Method - Natural
The researcher takes advantage of a naturally occurring IV . This is a natural experiment because the variable would have changed even if the experimenter was not interested
- not always natural, can be lab
Natural AO3
:) High ecological validity as natural experiments are often based in a real life setting therefore it is easier to generalise the finings beyond the setting of the study to other similar settings (CONTEXT) increasing the external validity.
:( Low control over extraneous variables as natural experiments usually take place in a natural environment (CONTEXT). Therefore, it is difficult to establish cause and effect between the IV and the DV lowering the internal validity of the study.
Experimental Method - Quasi
experiment which has an IV based on an existing individual difference between people (characteristic) which the researcher has not manipulated e.g. age or gender and the researcher measures the effect on the DV.
Quasi AO3
Quasi can be conducted in a natural or controlled environment (Field or Lab). therefore, depending on the scenario, you would have to select the appropriate evaluation (STRENGTHS AND WEAKNESSES
Problems with experimental method 1
Demand characteristics – Any clues given off within the research, this could lead to the Ppt’s change their natural behaviour choosing to either help or hinder the researcher (CONTEXT), reducing the internal validity
-controlled by: Getting another researcher who does not know the aims of the research (CONTEXT) to carry out the experiment. Therefore, they cannot give off clues to influence the ppts. (single/double blind)
Problems with experimental method 2
Bias in research - when the researcher is designing the resources/experiment or collecting results. Their own opinions/values guide decision making
Control: Randomisation.
Where possible aspects of the research should be random, not decided by the researcher which may be open to bias
1. all potential items go into hat
2. draw out as many needed for condition 1 (link)
3. draw out as many needed for condition 2 (link)
Standardisation
all participants experience the same environment and instructions, to avoid extraneous variables
HOW:
Using exactly the same formalised instructions and procedures for all ppts in a research study (CONTEXT).
This includes standardised instructions that are read out to all ppts / printed and given to the ppts to read. Matching task difficulties
Experimental design- Independent groups
Ppts take part in one condition only. Each condition has different group of Ppts.
How to conduct it:
*A group of ppts are recruited.
*They are then split into equal sized groups.
*A different group is allocated each condition.
Independent groups AO3
:) There will be no order effects as ppts only take part in one condition so they will not get bored/fatigued or better at the task/have lasting effects from one condition to the next (CONTEXT) UNLIKE repeated measures design where ppts take part in all conditions and could have order effects, lowering the internal validity.
:( Individual differences may affect the DV as there are separate groups of Ppts (CONTEXT) in each condition e.g., some Ppts could have a better memory which wouldn’t be suitable in a memory study (CONTEXT) lowering the internal validity UNLIKE repeated measures that uses the same group of individuals in all conditions to reduce the effect of individual differences.
Experimental design- Repeated Measures
All ppts take part in all/both conditions of the experiment.
How to conduct it:
*Ppts complete one condition.
*After a sufficient time lapse (an hour) they will complete the other condition.
*The task (DV) should be different but matched on difficulty/measuring the same thing.
Repeated Measures AO3
:) Individual differences between participants in each condition are Link removed e.g., IQ as potential confounding variables, as the same Ppts take part in all conditions (CONTEXT). Therefore, the researcher can be surer that any conditions change in the DV are due to the IV and not any other variable increasing the internal validity of the research UNLIKE
IGD that will use different ppts in different conditions so individual differences could affect the DV.
:( Order effects may occur when participants take part in more than one experimental condition, they may perform worse in the second condition (due to boredom or fatigue) or better in the second condition (due to practice), or there may be lasting effects from condition 1 to condition 2 (CONTEXT) lowering the internal validity UNLIKE independent measures where ppts only take part in one condition so cannot be susceptible to order effects.
Experimental design- Matched pairs
Ppts are matched on important characteristics in the study e.g. age or IQ. By giving them an appropriate test and matching two ppts with the same score.
Each member of the pair is then randomly allocated across each condition
How to conduct it:
*Choose a variable to match ppt on
*Test the ppt on this variable before the study
*Match ppt with the same scores
*Randomly allocate each member of the pair to one condition
Matched pairs AO3
:) There will be no order effects as ppts only take part in one condition so they will not get bored/fatigued or better at the task/have lasting effects from one condition to the next (CONTEXT) UNLIKE repeated measures design where ppts take part in all conditions and could have order effects.
:( Very time consuming to match up and find similar participants. (CONTEXT) especially if a pre-test is needed to match ppt on a certain characteristic. The psychologists would need a very large pool of people UNLIKE repeated measures design where you use the same ppts in all conditions.
Problems with Independent Groups
Individual differences are a possible issue because there are different ppts in each condition, therefore in one condition the ppts may have a particular trait that makes them perform better at the DV (CONTEXT) – meaning it wasn’t the IV that affect the DV, lowering the internal validity.
Control: Random allocation
WHY:
This ensures that each Ppt has the same chance of being in one condition as any other condition, therefore it is unlikely that all ppts in one condition will share a characteristic to help them perform better at the DV.
HOW:
1. Write the names of the ppts on equal sized pieces of paper.
2. Put the paper in a hat
3. Then randomly pick out ppts names an place half in condition 1 and the other half in condition 2 (CONTEXT).
Problems with Repeated Measure
Order effects are a possible issue. Ppts take part in all conditions and so may perform better in the second condition (due to practice) or worse in the second condition (due to fatigue), or the 1st condition may have lasting effects on the 2nd condition.
Control: Counterbalancing (ABBA) method.
WHY:
To control and reduce order effects
HOW:
1. Half the Ppts complete condition A then condition B. (CONTEXT)
2. The other half complete condition B first then condition A.
3. Order effects are balanced
Sample Vs Population
The group of people who are the focus of the researcher’s interest are called the population
Whereas
A sample is a group of people taken from the target population to take part in the research. It is assumed that the sample is representative of the target population.
Sampling Technique- Random Sampling
Every person in the target population has an equal chance of being selected.
How?
Create a complete list of all the target population.
Write names on equal size pieces of paper and put in a hat.
Without looking select the number you want in your sample from the hat.
Random Sampling AO3
:) Sample will potentially be more representative of the target pop as everyone has an equal chance of being selected allowing the researcher to draw conclusions (CONTEXT) increasing the external validity.
Discussion- However there is still chance that an unrepresentative sample may be drawn.
:( Very time consuming because it takes time and can be difficult collecting all the names of participants to take part in the study (CONTEXT) and it is not guaranteed participants would want to take part in the research.
Sampling Technique- Systematic Sample
A ‘system’ is applied to select Ppts.
How?
Create a sampling frame to organise the target population e.g. put all the people in the target population in alphabetical order.
Decide on the system e.g. every 5th or 10th person is chosen until there are enough ppts for the sample.
Systematic Sample AO3
:) Avoids researcher bias as the researcher has no influence over who is chosen to take part in the study (CONTEXT) as it is objective once the system to select ppt is established.
:( The sample may not be representative as not everyone has an equal chance of being picked therefore it is difficult to generalise to the target population (CONTEXT). Moreover, creating the system and collecting names could be time consuming and difficult for the researcher.
Sampling Technique- Stratified Sample
An advanced way of sampling where the make-up of the sample reflects the proportions of certain sub-groups in the target population. E.g. is 70% of the target population are female then 70% of the sample will be female as well.
How?
1. Identify the different sub-groups that make up a population e.g. males/females.
2. Calculate how many of each proportion you would need for your sample size e.g. if you wanted a sample of 80 ppts what would 70% of 80 be.
3. No. of Ppts which make up each proportion are chosen using random sampling.
If the scenario does give you percentages:
(Sample size wanted/100) x Percentage size of sub-group
If the scenario doesn’t give you percentages: (Sample size wanted / Target pop size) x Sub-group size
Stratified Sample AO3
:) Most representative sample as is it accurately reflects the make-up of the target population (CONTEXT) in the correct proportions/amounts meaning findings can be generalised to the target population. Stretch: however, stratified sampling is not perfect. The strata cannot reflect all the ways people are different so complete representation of the target population is not always possible.
:( Very time consuming because it takes time working out the proportions of Ppts needed (CONTEXT) and it is not guaranteed participants would consent to take part.
Sampling Technique- Opportunity Sample
Being there at that moment in time – for example waiting for participants in the canteen during period 3 and asking the first 30 ppts who walk past to participate.
How?
Go somewhere where you will have access to the population you want to study and ask the first 30 ppts who are there at that moment in time if they will take part.
Opportunity Sample AO3
:) Very quick, easy and less money is required as you choose who is available at the time for your sample (CONTEXT) unlike random sampling which takes a long time to identify everyone in the target population and make a list of all names.
:( May not be representative of the target population, those picked by an opportunity sample are present in the same place at the same time, doing similar things, (CONTEXT) therefore they may have similar personalities (E.G), making it difficult to generalise findings to wider population (link to target pop). Lowering EV.
Sampling Technique- Volunteer Sample
This is ‘self selected’ method where people volunteer themselves to be participants, E.g. placing an advert for participants then waiting for the first 30 replies.
How?
Place an advert where people in the target population will see it and wait for the first 30 people to reply.
Volunteer Sample AO3
:) Very quick, easy and cheap to conduct as all that is required is to place an advert for the study in a place the target population may be (CONTEXT) and wait for ppt to volunteer to take part in the study (number) unlike random sampling which takes time to place names into a hat and give the population an equal chance of being picked.
:( The sample maybe bias/not representative as only a certain type of personality (extrovert/curious/more likely to please the researcher) will volunteer (CONTEXT) therefore difficult to generalise to the target population. Lowers EV.
Reliability definition
Reliability refers to the ability to repeat a study in similar conditions to gain consistent results.
Ways to assess/improve reliability
Test re-test:
1. Participants are given a task or measure to complete (CONTEXT)
2. The same participants are then given the same task (CONTEXT) after a time delay e.g. two weeks.
3. Correlate the results from each test (CONTEXT) using stats test.
4. A strong positive correlation of above +0.8 shows high reliability
Improve- operationalise
Types of external validity
Ecological validity- A form of external validity. The extent to which findings can be generalised beyond the setting of the study to other real life settings.
Population validity- A form of external validity. The extent to which findings can be generalised beyond the sample studied to the target population.
Temporal validity- A form of external validity. The extent to which findings remain true over time and can be generalised to other time periods.
Face validity (internal validity)
An independent psychologist in the same field looks at the experimental conditions (questions – if a questionnaire, behaviour categories – if an observation) to see if they look like they measure what they intend to measure (CONTEXT). If the researcher says ‘yes’ then the research is said to have face validity
Concurrent validity (internal validity)
A way of assessing validity by comparing the results of the new test (CONTEXT) with the results from another similar test which has already been established for its validity, using a stats test. If results from both tests (CONTEXT) are similar then we can assume the test is valid. The correlation of the two results gained from an appropriate stats test should exceed +0.8.
Improving Validity (internal validity)
Experimental Research
- Use of a control group to assess that the IV has affected the DV (establishing cause & effect).
- Standardised procedures/instructions to reduce investigator effects.
- The use of single blind procedures to reduce demand characteristics & double blind procedures to reduce demand characteristics and investigator effects.
What is meant by a pilot study?
A pilot study is a small scale trial run of the research before the real main scale research takes place to identify any problems and AMEND them.
What are the aims of a pilot study?
- Check standardised procedures and general design
- Check the amount of time given to ppts is enough to complete the task (LINK TO THE SCENARIO, WHAT IS THE TASK??)
- For questionnaire or interview to make sure the questions are clear
- For observations check coding systems and behaviour categories are understood, check observer/camera in a good position to see behaviours.
- Extraneous variables can be identified and possibly eliminated (LINK - GIVE AN EXAMPLE)
- Ask the ppts to discuss their experiences of the experiment
- ANY PROBLEMS CAN BE IDENTIFIED AND ADJUSTED BEFORE THE MAIN STUDY
↑ I.V / save money and time
What are ethical issues?
issues that participants could face during the research that can affect them in many ways.
Ethical issues arise when there is conflict or dilemma exists between the rights of the Ppts and the researchers need to get valuable and meaningful findings.
Ethical issues
Deception- Deception is when the ppts are not told the true AIM of the study deliberately, to prevent participants from showing demand characteristics.
Informed Consent- Where ppts give permission to take part in the study after being informed of the true aims of the research, as well as being given the right to withdraw.
Protection of Participants- Ppts should be protected psychologically and physically. They must not be placed at more risk than they would be in everyday life.
Confidentiality- Ensuring ppts details and data (findings) remain anonymous so they cannot be identified in the research. The researcher could use fake (pseudo names) names/initials e.g. Genie and Patient KF.
How can we deal with ethical issues in research? (British psychological society)
Deception:
Debrief - a post research interview where the Ppt is told the true aims of the research and full details of the study are given e.g. knowledge of any other experimental groups.
- Ppts should also be given the right to withdraw their data from publication.
- Ppts must be reminded that their behaviour is typical/normal of the general public, and in extreme cases if ppts feel embarrassed or stressed as a result of the procedures used in the research, they should be offered counselling.
Informed consent:
Participants should be given a consent form detailing the true nature and aims of the study.
A child under 16 cannot give their consent.
There are other ways to gain consent - the 3 P’s
1. Presumptive consent - this is gaining consent from a similar set of ppts who won’t actually be doing the study.
2. Prior informed consent – getting general consent from ppts of things that ‘could’ happen.
3. Parental consent – when ppts are too young to give consent this is gained from parents. (under 16)
If consent wasn’t gained from the Ppts in the study a debrief should be given and Ppts can give retrospective consent. Ppts may wish to withdraw their data if they do not consent.
Protection from harm Right to withdraw:
Ppts should be offered the right to withdraw at any point during or after the research has taken place.
Debrief - reassure the Ppts that their behaviour is normal - offer counselling to all Ppts.
Confidentiality:
Anonymity - All personal should be kept anonymous e.g. use of numbers, initials or fake names.
