Research Methods Flashcards

Question

What is systematic sampling?

Answer 1

Get all the names of target population and order them in some way. Every nth member of target population is selected. This creates a sampling frame.

Answer 2

+ unbiased selection (researcher isn’t choosing the sampling frame). + should be representative so can generalise. - may still be unrepresentative so can’t generalise. - time-consuming.

Answer 3

Identify subgroups that are relevant to your research. Work out the proportion of that subgroup in the target population. Make sure the proportion is reflected in the sample. Choose participants with random sampling.

Answer 4

+ will be representative, so can generalise. + no researcher bias. - time-consuming. - only representative of the chosen subgroup.

Answer 5

Use people most convenient and available to you that fit the criteria (means people are similar though).

Answer 6

+ not time-consuming. - unrepresentative so cannot generalise. - researcher chooses sample so could result in bias.

Answer 7

Participants are self-selected. Researcher advertises the research, often with a reward.

Answer 8

+ not time-consuming. - unrepresentative as certain people more likely to volunteer, so can’t generalise. - potential for no volunteers.

Answer 9

A scientific process involving the manipulation of the IV to determine cause and effect. It must be ensured that only the IV has an impact on the DV.

Answer 10

In a well-controlled environment, where accurate measurements are possible. Standardised procedure (all participants experience exact same). IV manipulated by researcher. DV measured. Participants randomly allocated.

Answer 11

In everyday environment of participant (not controlled). Less control of extraneous variables. IV manipulated by researcher. DV measured.

Answer 12

No control over IV as it occurs naturally in real life. Measures DV (could be secondary data though). IV in natural setting but DV could be measured in lab/field. Often conducted when it’s unethical to manipulate IV.

Answer 13

Naturally occurring IV. The IV is difference between people that already exists (personal characteristic eg. gender, age) DV measured (could be in lab/field).

Answer 14

The degree to which the materials and procedures involved in an experiment are similar to events that occur in the real world. Type of external validity.

Answer 15

Different levels of the IV.

Answer 16

How participants are allocated to the different conditions in an experiment.

Answer 17

Different participants are used in each condition of the IV. Each participant only does 1 condition. This means each condition of the experiment includes a different group of participants.

Answer 18

The same participants take part in each condition of the IV. Each participant does all conditions. This means each condition of the experiment includes the same group of participants.

Answer 19

Each condition uses different but similar (matched) participants. An effort is made to match the participants in each condition in terms of any important characteristic which might affect performance (DV). Each participant only does 1 condition.

Answer 20

Identify an appropriate variable (which might affect the DV). Eg. IQ. Pre-test the variable eg. IQ test. Pair up most similar participants eg. pair up with same/similar IQ score. Randomly allocate 1 to each condition eg. split pair into each condition.

Answer 21

Test-retest (experiments, questionnaires, structured interviews - standardised procedure) participant does test, same participant does same test again at later event/date. Researcher does Spearman’s Rho correlation with both results - 80% is reliable. Interrater/interobserver (observations, content analyses) 2 researchers agree on behaviour categories. Both researchers observe at same time with same behaviour categories. Each researcher tallies independently. Spearman’s Rho correlation with both results - 80% is reliable.

Answer 22

At face value - researcher considers if it looks like a good measure, may ask experts or participants too. Concurrent validity - compare test to an established test. Participants do both tests. Then Spearman’s Rho correlation of both results - 80% is reliable.

Answer 23

Population validity - if results can be generalised beyond the population. Ecological validity - if results can be generalised beyond setting/situation. Temporal validity - if results can be generalised across time. Cross-cultural validity - if results can be generalised across cultures.

Answer 24

When the differing characteristics between participants in conditions may impact the DV.

Answer 25

When a participant does more than one condition and as a result becomes better (practice effects) or worse (fatigue effects) over time.

Answer 26

No order effects - participants only take part in one condition. Less chance of demand characteristics - participants are less likely to guess the aim.

Answer 27

Individual differences - may impact DV. Need twice the amount of participants.

Answer 28

Choose conditions based on chance (random allocation). This should distribute characteristics (in theory).

Answer 29

No individual differences (won’t impact DV). Requires less participants.

Answer 30

Order effects - as participants participate in more than one condition. Risk of demand characteristics - participants are more likely to guess the aim.

Answer 31

Participants in each group participate in conditions in different orders (counterbalancing). Eg half do A then B, other half do B then A. Order effects still occur, but balance each other out as they occur equally in both groups.

Answer 32

Less individual differences - less likely to impact. No order effects - participants only take part in one condition. Less chance of demand characteristics - participants less likely to guess the aim.

Answer 33

Time consuming - can be difficult to find participants similar enough to pair, pre-test etc.

Answer 34

Conditions are assigned to participants through chance (eg using a random generator).

Answer 35

The first group does condition A then B and the second group does condition B then A.

Answer 36

Any variable (other than IV) that, if not controlled, could have affected the DV. This reduces validity. They only influence the DV so don’t vary systematically with the IV. Either participant or situational.

Answer 37

Individual differences between participants that might affect the DV eg age, intelligence, gender, personality.

Answer 38

Features of the environment/research situation which might affect the DV eg noise, weather, instructions etc. Easier to control as a researcher than participant variables.

Answer 39

Any variable (other than IV) that, if not controlled, could have affected the DV. Vary systematically with the IV (when IV changes, so does CV) so associated with both DV and IV. Very difficult to control DV so can’t be sure if it’s the IV or the CV that’s affecting the DV - difficult to establish cause and effect.

Answer 40

When the researcher consciously or unconsciously impacts the DV - therefore they’re acting as en extraneous variable. This could be due to: Their knowledge of what the research aim and hypothesis is (may subconsciously or unconsciously change their behaviour in order to accept this). Their traits (personally characteristics that may influence participant behaviour - unconscious).

Answer 41

When the participant consciously or unconsciously changes their behaviour (becomes unnatural behaviour) and so impacts the DV and acts as an extraneous variable.

Answer 42

When participants work out the aim of the study and play along to help the researcher. Tends to only impact one condition.

Answer 43

When participants work out the aim of the study and do the opposite to hinder the researcher. Tends to only impact one condition.

Answer 44

When participants act differently as they want to act in the most socially desirable way to look good. Tends to impact across all conditions. In general, prevalent in observational studies/questionnaires.

Answer 45

It allows researchers to better establish cause and effect. This improves internal validity.

Answer 46

Using standardised procedure, meaning all participants experience the research in the exact same way (apart from the IV). Controls for situational variables.

Answer 47

The use of chance in research. Aspects of research are decided by chance to reduce the role of the researcher and so reduce investigator effects. Eg designing of materials, allocation of participants (random allocation) to conditions or the order of conditions - reduces participant variables.

Answer 48

The correct rules of conduct necessary when carrying out research. Researchers have a moral responsibility to protect participants from harm by protecting the participants. It also protects the reputation of psychology and the psychologists themselves.

Answer 49

When there’s a conflict between the rights of participants and the needs of researchers to conduct research that makes a meaningful contribution to society.

Answer 50

Respect - psychologists should respect the dignity of all people. This includes issues such as privacy, confidentiality, consent and deception. Competence - psychologists should have an appropriate ability to provide specific services to a professional standard. Responsibility - psychologists should accept responsibility for what is in their power, control or management. Psychologists should protect their participants from harm when conducting research. Integrity - psychologists should be honest, truthful, accurate and consistent in their actions, decisions, methods and outcomes.

Answer 51

Participants need to be informed of what will happen in the study, and what they will be expected to do, particularly if there is risk involved. Participants should be informed of anything in the study that the researcher thinks they may object to. They should also be told that they have the right to withdraw at any time Informed consent must be voluntary, and rational. Participants must be given relevant details to make an informed decision, including the purpose, procedures, risks, and benefits. And participants must have the capacity to rationally weigh up the decision. Eg children and people with learning disabilities can’t consent (parents can on behalf).

Answer 52

As well as fully informed consent there are 3 other options: Retrospective consent - participants are asked after the research. Presumptive consent - where a group similar to the participants is asked if they would find this acceptable. General prior consent - where participants consent but do not know all the details. If this is the case, the participants must be fully informed during the debrief (after the research has taken place) and they can have their results/contribution removed if they wish.

Answer 53

Sometimes it is necessary to deceive participants. This is because if participants knew the full aims of the study, they may alter their behaviour, making the results less valid (demand characteristics, screw-you-effect, social desirability bias). Frequently, in research, the researcher withholds information from the participants, rather than giving them false information. Deception is unethical as participants have the right not be lied to. If information about the aims of the study is withheld, they are unable to give informed consent to take part. This could lead to distress as the participants do not know what they have volunteered for. However often deception in experiments is minor and causes no harm.

Answer 54

Privacy is the right to control information about yourself. Also refers to the area of research. If our privacy is invaded, even with our consent, we still have the right to confidentiality. Confidentiality refers to the protection of the participants’ personal data. This right is enshrined in law by the Data Protection Act (1986). It is vital that whenever possible, data is kept anonymous so individual participants cannot be identified. However, with some research methods (eg case studies) this can be difficult.

Answer 55

Participants should be debriefed following participation in research (fixes deception, could gain retrospective consent). This becomes even more important in situations when they have not been aware they have been taking part in a study or where they have been deceived/not given fully informed consent. It is also important when they were not aware that there were PPs in another. Lack of debriefing could be unethical as it could fail to deal with other ethical issues.

Answer 56

Participants should have the right to withdraw themselves from the study at any time without consequence. Participants should be able to leave a study at any time if they feel uncomfortable. They should also be allowed to withdraw their data. They should be told at the start of the study that they have the right to withdraw. They may withdraw for any reason but particularly where they feel uncomfortable or distressed about what they’re being asked to do (must be paid full amount of money if money was offered for being a participant). The researcher’s behaviour should make participants feel like they are able to leave. It is particularly important that participants have the right to leave the study if they have been deceived, or information about the study withheld so that they are unable to give informed consent. When participants withdraw from a study, it can cause problems for researchers as it will bias the results, as the participants who stayed will be different from the ones who left.

Answer 57

Harm could be physical (such as being made to drink, smoke or exercise excessively) or psychological (such as stress, anxiety or embarrassment). Studies should be designed to ensure that participants come to no more harm than they would in everyday life. However, it may be that harm doesn’t become apparent until during or after the research and which can’t be anticipated. Researchers need to weigh up the importance of the findings of the research with the potential harm to participants. Any potential harm needs to be communicated to the pts so that they can give informed consent. Participants should leave a study in the same state/condition that they entered it.

Answer 58

1. Making sure participants understand the nature/aim of the study - hypothesis etc. Use clear terms/no jargon. 2. Giving participants the right to withdraw their results/data from the study (video/recordings). 3. Explaining any deception & why there was a need for it. 4. Explaining any conditions that the PPs were unaware of. 5. Checking participants for negative effects or misunderstandings. 6. Use counselling if necessary to ensure participants leave in a healthy state of mind. 7. Opportunity for questions/follow up - offer access to finished research. 8. Thanking participants for taking part in the research - giving up their time. It should also be written verbatim.

Answer 59

A cost-benefit analysis is used to decide if the benefit to society is greater than the cost of breaking ethical guidelines (which are then dealt with as effectively as possible). This means that sometimes we do break ethical guidelines.

Answer 60

Get participants to sign a consent form that’s detailed and contains all the necessary information. If deceived, they could get general prior consent, or during debrief, gain retrospective consent.

Answer 61

Inform participants before and after the study. Ensure this won’t cause psychological harm. Deception could be kept to a minimum - all deception should be necessary.

Answer 62

Don’t use names or specific descriptions of participants that would allow them to be identified.

Answer 63

Should be told before, throughout and after the study. Researcher’s behaviour should make participants feel like they can withdraw.

Answer 64

Communicate to participants that they may come to harm so they can give full informed consent. Ensure they come to no more harm than they would normally in everyday life. Do a cost-benefit analysis. Debrief afterwards, potentially even counselling.

Answer 65

They involve watching and recording people’s behaviour. They don’t have an IV.

Answer 66

Type of observation - researcher controls setting and situation. Involves the recording of spontaneously occurring behaviour but under conditions constructed by the researcher. The conditions are carefully controlled.

Answer 67

Type of observation - participants do day-to-day tasks. Involves observing spontaneously occurring behaviour in the participant’s natural environment. The behaviour they observe is less artificial and people usually have more choice in how they behave. They are usually done to observe behaviours that couldn’t be studied in a fully controlled setting.

Answer 68

Control of extraneous variables. Standardisation means it can be repeated and can check reliability. Can set up a situation where you’ll see relevant behaviour.

Answer 69

Lacks external validity (mainly ecological validity).

Answer 70

Is ecologically valid.

Answer 71

Less control of extraneous variables. Can’t be repeated due to lack of standardisation, so can’t check reliability. Can’t set up a situation where you’ll see relevant behaviour.

Answer 72

Type of observation. Where the psychologist observed an individual, group or situation without people being aware. The psychologist may pretend to be part of a group without revealing their true identity or aim. Alternatively, a psychologist may observe from a hidden viewpoint.

Answer 73

Can’t show demand characteristics as they’re unaware that they’re being observed. No social desirability bias.

Answer 74

Unethical - can’t give informed consent, could involve deception.

Answer 75

Type of observation. When the psychologist is open about their observation. They make their presence obvious and people know that their behaviour is being recorded.

Answer 76

Ethical - can give informed consent.

Answer 77

Participants may show demand characteristics. May shows social desirability bias.

Answer 78

Type of observation. Where the observer joins the group that they’re observing and becomes part of it.

Answer 79

Likely to gain more information as they won’t miss anything.

Answer 80

Researcher bias. Less internal validity.

Answer 81

Type of observation. Where the psychologist is not directly involved in what is being observed and records data from a distance.

Answer 82

No researcher bias - researcher won’t impact behaviour. Higher internal validity.

Answer 83

Likely to miss information as they’re observing at a distance.

Answer 84

Type of observational design. The researcher will write down/record everything they observe.

Answer 85

Learn a lot of information (qualitative data) - detailed understanding.

Answer 86

May record irrelevant information. Subjective to the researcher. Complex to analyse the data. Not a good technique to use in busy areas - may miss details/information.

Answer 87

Type of observational design. The researcher will identify target behaviours which they’re interested in. These target behaviours will then be broken down into behaviour categories and only these behaviours will be recorded. Any behaviour that isn’t a behavioural category will not be recorded and will be ignored. The researcher will then use a specific sampling method to decide how often they record the behavioural categories.

Answer 88

More objective - improves validity and reliability. Researcher is less likely to miss behaviour. All behaviour is relevant to the aim.

Answer 89

Don’t learn much about the behaviours due to quantitative data.

Answer 90

Once a target behaviour is identified, it’s broken down into precisely defined components which are observable and measurable. They must be objective and specific. It’s important that the researcher ensures all the ways in which target behaviour may occur in their behaviour categories otherwise they will miss relevant information.

Answer 91

When you’ll observe behaviour. IN OBSERVATIONS.

Answer 92

Used in unstructured observations. All behaviours observed during a period are recorded. For the whole observation. Eg if time frame is 1 hour, record all behaviours in that hour.

Answer 93

Produces detailed information.

Answer 94

May not be possible to record all information.

Answer 95

Used in structured observations. Time is split up into intervals (eg every 30 seconds) and at every interval a recording of the behaviour taking place is made.

Answer 96

More manageable and less likely to miss behaviour categories at the specific times.

Answer 97

May miss behaviour categories if they aren’t observed at the specific time.

Answer 98

Used in structured observations. Observation are made of a specific event whenever if occurs. Eg if target behaviour is fighting, then each fight will be observed but nothing else. It avoids the collection of irrelevant behaviour.

Answer 99

Avoids the collection of irrelevant data. Will record all behaviour categories so if they’re infrequent, they won’t be missed.

Answer 100

Less manageable and may miss behavioural categories if they’re very frequent/many are observed at the same time.

Answer 101

Naturalistic observations are set in real-life situations, so they tell us about real/natural behaviour as the setting is not artificial (high ecological validity). People have less chance of working out the aim, especially if the observation is covert. Findings are more reliable because the researcher can see for themselves how participants behave rather than relying on self-reports. Inter-observer reliability - where two (or more) researchers carry out the observation at the same time. The two (or more) sets of data are analysed using a correlational analysis to see the similarity (or not) between the observations. Observational studies are likely to be more holistic and less reductionist than is often the case with experimental studies. It is observations of total behaviour rather than with small elements in isolation. Observations can provide hypotheses for more searching examination and experimentation.

Answer 102

Loss of control - there is no control over extraneous or confounding variables which can affect the results which lowers the validity of the findings and conclusions. Causality - It is harder to imply cause in observational studies. (e.g. Hargreaves observed that boys in lower sets had poorer attitudes towards school and lower attainment than boys in the higher sets. However, this may be due to streaming, or socio-economic conditions, or intelligence etc.) Observer bias can be a problem - if you are looking for a particular behaviour you are more likely to find it. It is difficult to remain dispassionate and neutral (can be minimised through inter-observer reliability). If participants are aware they are being observed then they may act differently, giving invalid results. This is known as the observer effect. The ecological validity of the behaviour is reduced as the situation is no longer natural. Difficult to make judgments about thoughts and feelings as these features are not clearly observable. There may be ethical problems with consent and the invasion of privacy with covert observations.

Answer 103

A measure of the extent to which 2 variables are related. In a correlation, the 2 co-variables are each measured and the relationship between the 2 is established. There is no manipulation or naturally occurring change so there’s no IV and therefore no DV for it to have an effect on in correlations. In a correlation, we don’t know what variable is impacting the other, or if there’s a 3rd variable impacting both - we just know that the 2 variables change systematically with each other.

Answer 104

When the 2 co-variables both increase or decrease together (both move in same direction).

Answer 105

When the 2 co-variables move in opposite directions (as one increases, the other decreases).

Answer 106

When there’s no relationship between the 2 co-variables.

Answer 107

Different to an experimental hypothesis as it doesn’t predict how one variable will affect another, instead it predicts if there will be a relationship between the 2 co-variables.

Answer 108

Scattergrams are used to display correlations. The x axis represents 1 co-variable and the y axis represents the other. Each cross will represent 1 participant. Each pair of co-variables is used to plot 1 data point on the scattergram. This is repeated for each pair of co-variables. The pattern of the data points indicate the direction of the correlation between the co-variables. A line of best fit can be drawn to help identify the strength of the correlation.

Answer 109

Identified how strong the relationship is between 2 co-variables. Calculated using a statistical test. A correlational coefficient is always a number between -1 and +1. It’s represented by a lower-case r (eg r=-1). Most correlation coefficients have a + or - sign, which shows whether it’s a positive or negative correlation. A correlation coefficient of +1 represents a perfect positive correlation, and a correlation of -1 is a perfect negative. A correlation coefficient of 0 represents a zero correlation. The closer the correlation coefficient is to +/-1, the stronger the correlation.

Answer 110

The co-variables increase together up to a certain point and then as one increases the other decreases, or vice versa. Looks like an upside-down V.

Answer 111

Allows a researcher to establish that 2 variables are related in some way (can identify a trend within a data set and are often used to prompt further research in an area). If the correlation coefficient isn’t significant then a causal relationship can also be ruled out (don’t waste time and money in further research in the area). Provide a precise measure of how 2 variables are related (quantitative data). Means results are objective. Usually easily repeatable, so reliability can be checked. Relatively quick compared to other methods as researcher only measures and can use secondary data. Allows researchers to statistically analyse naturally occurring behaviour which may not be able to study experimentally due to practical or ethical reasons eg impact of stress levels.

Answer 112

No cause and effect - correlations only tell us the relationship/association between 2 variables, not that one caused the other - there may be several variables that are interrelated which are called intervening variables. Complex relationships are hard to measure eg in curvilinear relationships, a correlation coefficient would likely be misleading and very simplistic.

Answer 113

A technique for collecting information from many people by asking them questions. Involves asking people to report their own thoughts, feelings and behaviours. There are 2 major methods used in self-reports: questionnaires and interviews.

Answer 114

A research instrument consisting of a series of questions for the purpose of gathering information from respondents. They can be carried out face-to-face, by telephone, computer or post.

Answer 115

No restrictions on participants’ answers. Allows the participant to respond in their own words and usually will lead to an extended response. This would produce qualitative data.

Answer 116

Participants choose from a fixed set of responses determined by the researcher. These could be as simple as yes/no or a number of fixed choices or a scale. This would produce quantitative data.

Answer 117

Rating scale questions - where responses are on a numerical scale and each value has a meaning which usually represents their strength of feeling toward a topic. Likert scale questions - where participants answer on a scale of how much they agree/disagree with a statement. Fixed choice options - contain a list of possible options and respondents tick any items that apply to them.

Answer 118

Data is more in-depth so provides a deeper understanding of behaviour/thoughts/feelings.

Answer 119

The data is qualitative which makes it harder to analyse. Provides qualitative data which is subjective and unscientific.

Answer 120

The data is quantitative which means that it’s easier to analyse. Provides quantitative data which is objective and scientific. Answers will always be relevant to the question asked.

Answer 121

Data lacks detail and doesn’t explain behaviour/thoughts/feelings.

Answer 122

Aims - are all questions relevant to the aims of the research? Length - shouldn’t be too long (participants won’t want to complete it) or too short (you don’t acquire enough information). Pilot study - run a small-scale practice study to ensure people understand the questions. Question order - questions should be in a logical and progressive order. Terminology - no jargon and appropriate for target audience. Presentation - professional (ensure participants take it seriously). Question phrasing - written with clarity, no leading questions, double-barrelled questions or double negatives.

Answer 123

Social desirability - unlikely to occur as people are anonymous. Sample size - can be sent out to a large sample and so obtain large amounts of data. Efficiency - once designed, the questionnaire takes minimal time and resources for researchers. Replicability - same questionnaire can be used again. Ethics - no manipulation and completion is optional.

Answer 124

Social desirability - it’s easy for participants to lie in a questionnaire to make themselves look better (reduces internal validity). Response rate - may get a limited number of responses, certain people are more likely to respond (response bias), both of previous issues reduce questionnaires from being representative.

Answer 125

Different from questionnaires as they involve social interaction. Involves direct verbal questioning of participants by the researcher (1:1 or group setting). Typically done face-to-face or over the phone.

Answer 126

Structured interview - questions asked in a set/standardised order and the interviewer will not deviate from the interview schedule or probe beyond the answers received (not flexible) and would write down everything they answered with. These are based on structured, close-ended questions. Semi-structured interview - some pre-set questions. More flexible, allows new ideas to be brought up during the interview as a result of what the interviewee says. Unstructured interview - an interview schedule may not be used and even if one is used, they’ll contain open questions that can be asked in any order (complete flexibility). Some questions may be added/missed as the interview progresses.

Answer 127

Replicable due to standardised questions, so can check reliability. Researcher bias is reduced as they’re asking all the same questions in the same order.

Answer 128

May miss important information as researchers can’t deviate from the predetermined questions and so can’t follow-up interesting areas. Predetermined questions may restrict answers which may force people to give an answer that doesn’t fully represent their view.

Answer 129

TO AN EXTENT. Flexible and allows detailed data collection which can be tailored to areas of interest. Allows for a greater understanding. More natural so participants are more likely to give natural responses.

Answer 130

TO AN EXTENT. Unstandardised so not replicable, making it hard to check for reliability.

Answer 131

Flexible and allows detailed data collection which can be tailored to areas of interest. Allows for a greater understanding. More natural so participants are more likely to give natural responses.

Answer 132

Difficult to analyse and draw conclusions due to the large and varied amount of data collected. Unstandardised so not replicable, making it hard to check for reliability.

Answer 133

Interview schedule - a set of prepared questions designed to be asked exactly as worded (structured and semi-structured interviews). Recording/analysing responses - how information will be recorded (writing answers, voice recorder, mind-mapping). Take special care with vulnerable groups - adapt the interview for children (not complex language, short length, familiar environment, toys) or elderly (comfortable environment, short length, speak loudly). Language in questioning - consider use of terminology. Sensitive topics - identify potential topics, adapt as necessary (abuse, mental health). Investigator effects - ways in which the researcher may impact the participants’ response (wording of questions, mannerisms, age, gender, personality traits).

Answer 134

More flexible - interviews allow the researcher to explore complex issues (eg trauma) which may be difficult to do using other methods. Rich/detailed information - during interviews, you’re likely to gather detailed information about how people think/feel as qualitative data is collected. This gives more insight or greater understanding of a topic/behaviour. Less social desirability - people may be less likely to lie than when during a questionnaire (not anonymous).

Answer 135

Interviewer bias - expectations (bias) of the interviewer may influence the interviewees’ responses and lead them to give an answer that is not what they really think or feel (or researcher may interpret data in favour of their aim). Efficiency - can be significantly more time and resource consuming than questionnaires (sometimes 1:1, have to record answers/transcribe them).

Answer 136

Provides detailed and in-depth analysis of an individual, group, or event. May involve analysis of unusual individuals or events. Uses a range of methods to gather data (could include interviews, questionnaires, secondary data, observations, small experiments). Usually longitudinal. Collects qualitative data.

Answer 137

Detail - able to learn about the case in depth as it’s longitudinal and a wide range of methods are used. Practical/ethical - no manipulation is required - allows us to measure things we couldn’t practically/ethically study with other methods. Generate hypotheses - findings give us new ideas and we can then go out and test if it’s true for the population.

Answer 138

Subjectivity - open to interpretation due to qualitative data - the researcher could analyse it in a way that fits their aim. Generalisability - case studies aren’t representative as they’re of unusual individuals, so results can’t be generalised. Validity of self-report - social desirability - sometimes people may lie or the answers they give (may not always be relevant to the aim) so reduces internal validity. Inefficient - longitudinal. Researcher bias - likely to occur as it’s longitudinal and researcher is involved - likely to influence behaviour of participant, may interpret data in a biased way. Can’t establish cause and effect - have no level of control, many extraneous variables. Often happen after the event has happened - don’t have a before and after.

Answer 139

Numerical. Represents how much or how long, how many etc. there are of something. The dependent variable in an experiment provides quantitative data. Closed questions in questionnaires & interviews collect quantitative data - numerical information about your age, how many hours you work in a week, how highly you rate different TV programmes. In an observational study, a tally of behavioural categories is quantitative.

Answer 140

More objective and less open to bias. Easier/quicker to analyse than qualitative data and see patterns. Statistical analysis can be carried out.

Answer 141

Can overlook broader themes and relationships by focusing too much on numbers and statistics (confirmation bias). May fail to reflect real life (lacks ecological validity.

Answer 142

Non-numerical. It can be used to understand how an individual subjectively perceives and gives meaning to their social reality. It cannot be counted or quantified, but it can be turned in quantitative data by placing the data into categories and then counting frequency. Open questions in questionnaires collect qualitative data. It is usually detailed and descriptive. In an observational study if researchers describe what they see, rather than count frequencies of behaviours this provides qualitative data.

Answer 143

Provides rich, detailed data (better understanding). Often provides the participants with more room to express their thoughts/feelings on a subject. More likely to provide insight into the participants’ world and tends to have greater levels of validity.

Answer 144

Often difficult to analyse in terms of patterns and comparisons. Conclusions are difficult to draw and often rely on the subjective interpretations of researchers. Less reliable - sometimes impossible to repeat over time, and across researchers (subjective).

Answer 145

Original data that’s been collected specifically for the purpose of the investigation by the researcher. It’s data that arrives first-hand from the participants themselves. It can be collected via an interview, questionnaire, experiment or observation.

Answer 146

Fits the purpose - relevant as it meets the researcher’s needs.

Answer 147

Requires time, expertise, participants and money - impractical.

Answer 148

Data that already exists. It’s been collected by someone else for their purpose/research question. Desk research.

Answer 149

Usually inexpensive and quick to gather.

Answer 150

Cannot always be certain of the accuracy/could be outdated. May not be fit for purpose - May not answer the researcher’s particular research question.

Answer 151

A technique for combining the findings from numerous studies of a certain research area. Both quantitative and qualitative data can be analysed using meta-analysis. If the researcher only discusses the findings, this is qualitative data. If the researcher uses a quantitative approach, they’ll perform a statistical analysis of the combined data. Results from different studies are used and a general conclusion is produced. No additional research is carried out, data from previously conducted research is collected and analysed.

Answer 152

Allows us to view data with much more confidence. This is because they draw conclusions based on evidence from multiple empirical sources. Therefore, there’s an increased likelihood that meta-analysis findings will be more valid than independent experimental research that forms a conclusion based on a single study’s findings. Relatively cheap, as the studies have already been conducted and the results are already available. Meta-analysis in research has many practical applications in psychology - eg can provide a reliable, precise summary of whether an intervention is effective as a treatment method.

Answer 153

Researchers need to ensure the research studies they’re combining into their meta-analysis are reliable and valid, as this can affect the reliability and validity of the meta-analysis. The studies included in the meta-analysis will likely use different research designs, raising the question of whether the data is comparable. Publication bias - file drawer problem, biased selection.

Answer 154

Categorical. Separate categories and within each category, there’s a frequency (count how many people are in that category). Nominal scales are used for labelling data that has no quantitative value.

Answer 155

Ranked/unstandardised scale. Ranked - put participants in an order and number them. Unstandardised scale - values may not be the same for each person. Space between each isn’t standardised.

Answer 156

Standardised scale. Measured using units of equal intervals - level of measurement is precise. We know the order and exact differences between each variable.