Research Methods Flashcards
1
Q
What is an experiment?
A
Where a researcher controls variables, avoids biases and remains objective to test an experimental hypothesis; to determine cause and effect.
- There is an IV manipulated by the researcher.
- The effects of the IV on another variable is observed or measured DV)
- The participants are allocated randomly to the conditions.
2
Q
What is Quantitative Data?
A
Quantity: Numbers/Measurements
3
Q
What is Qualitative Data?
A
Quality: Thoughts/Opinions/Attitudes
4
Q
What is Primary Data?
A
Collected directly by the researcher
5
Q
What is Secondary Data?
A
Researcher collects from other people’s research – second-hand.
6
Q
What are Levels of Measurement?
A
- Interval (data from a scale)
- Ordinal (rank ordered data)
- Nominal (individual categories)
7
Q
What is an hypothesis?
A
A testable statement of predicted outcome.
8
Q
What is operationalisation?
A
- Clear definitions of the variables are made in terms of how they can be measured.
E.g., females tend to perform at least one grade better than their male peers in their A-Level Psychology exams
9
Q
What is the Dependant Variable?
A
That which the researcher measures, so wording must indicate that numerical data is collected.
10
Q
What are Individiual Differences as extraneous variables?
A
Natural memory, left or right-handed (handedness), sex, age, natural athleticism (or any naturally occurring individual variable).
11
Q
What are Uncontrolled situational effects as extraneous variables?
A
Time of day, available sunlight, ect.
12
Q
How do demand characteristics occur?
A
- A cue that comes from the researcher or the research situation that may give away the purpose of the research when interpreted by participants – this may lead to a change of behaviour in participants during the
research. - Participant may behave in a way they believe is expected of them (to please the researcher, for example) or even behave in a way which sabotages the research by deliberately ‘underperforming’ (by the participants
standards).
13
Q
What is a directional hypothesis?
A
When the hypothesis suggests a direction in the predictive statement eg. Group A will score significantly higher than Group B.
14
Q
What is a non-directional hypothesis?
A
When the hypothesis suggests a difference between two groups in the predictive statement eg. There will be a difference in the scores between Group A and Group B.
15
Q
What is a null directional hypothesis?
A
A predictive statement which suggests the IV will have no effect on the measured outcome (DV) eg. The style of learning will not have an effect on measured recall (Note: not the opposite, but rather no effect)
16
Q
What are labratory experiments and their evaluation?
A
- Take place in an environment that is strictly controlled and where the behaviour does not normally occur (does not need to be a traditional science laboratory).
(+) Good control (IV’s, DV’s & EV’s), good for scientific measurements, testing hypo’s and good replication
(-) Low ecological validity due to artificial behaviour - demand characteristics. Also lacks of mundane realism
17
Q
How does natural environment research work?
A
- Research takes place where the behaviour naturally occurs, however, variables are sometimes difficult to scientifically control
18
Q
How does controlled environment research work?
A
Research takes place in a strictly controlled (laboratory type) environment where the behaviour does not naturally occur
19
Q
How do field experiments work and evaluate them?
A
- Experiments where the behaviour would naturally occur, but the variables only occur because the researcher has set them up
(+) High Ecological Validity, takes place in the real world also meaning less demand characteristics
(-) Hard to control, measure and replicate meaning more influence from confounding variables.
20
Q
What are Naturalistic experiments and evalaute them?
A
- Experiments with naturally occurring IV’s (eg Sex or those involved in a train crash) and usually take place in the ‘field’.
(+) High Ecological Validity, IV is natural as in the real world. Also less demand characteristics and can be used in situations where it could not be created due to ethical reasons.
(-) Hard to control especially who
takes part… self-selecting so biased, measure,
replicate, and problem with IV’s
(Naturally occurring so open to bias)
and can raise ethical issues…
21
Q
What is High Ecological Validity?
A
The experiment takes place in a natural environment, where it naturally occurs and therefore supports the validity of any research claims as it is generalisable.
22
Q
Whtat is Low Ecological Validity?
A
The experiment takes place in an unnatural environment, so it is potentially an unnatural behaviour and therefore reduces the validity of any research claims.
23
Q
What is External Validity?
A
The degree to which a research finding can
be generalised to other situations outside of
the research setting. There are three types of
external validity:
- Temporal
- Ecological
- Population
24
Q
What is Face Validity?
A
Simple way of assessing whether a test measures what it claims to measure which is concerned with face value – e.g. does an IQ test look like it tests intelligence.
25
What is Concurrent Validity?
- A measure in agreement with pre-existing
measure that are validated to test for
the same concept? This is gauged by
correlating the measures against each
other - e.g. comparing a new test with another test of the same thing to see if they produce similar results.
26
What is Random Sampling?
- All members of the target population have an equal chance of getting picked.
- An example of random sampling is the lottery method – a list of all members in the population is obtained. The names are put into a lottery barrel or hat and names are randomly chosen until the researcher has their required sample.
27
What is opportunity sampling?
- Researchers select anyone who is willing and available to take part in the research.
- Researchers take the chance to ask whoever is around at the time (e.g., asking students from around the university to take part in research).
28
What is Self-Selecting/Volunteer Sampling?
- Involves participants selecting themselves to be part of the sample – it can also be referred to as self-selection.
- Adverts may be placed in newspapers or around institutions. Alternatively, people may approach researchers themselves when they hear of the research (such as putting their hand up in a class when the research is being
discussed).
29
What is Independent Group Design?
- Participants are allocated to a different group in an experiment, where each group represents a condition of the study.
- If there are two levels of the IV, then participants will experience one level of the IV only. E.g., one group taking antihistamines (to measure their effect on drowsiness and speech) vs one group taking a placebo.
- The performance of the groups would then be measures and compared, e.g., the number of words spoken by participants who took antihistamines compared to those who took the placebo drug.
- **Random allocation** may be used in independent groups designs – this ensures that every participant has an equal chance of being in one condition as any other condition. This is an attempt to control for participant variables that may arise.
30
What is Stratified Sampling?
Where you match the internal dynamics of a population in proportion. EG. Research with
year 12 at a school [made up of 60 boys and 30 girls], so use a ratio of 6 boys to 3 girls in
that proportion.
31
What is Repeated Measures Design?
- All participants take part in all conditions of the experiment.
- E.g., participants take part in condition A (experimental condition), then take part in condition B (control condition).
- Repeated measures designs ensures that participants are being compared ‘like for like’, compared to independent group designs which assesses the performance of two (or more) different groups ofparticipants in two (or more) conditions.
- **Counterbalancing** may be used in repeated measures designs to control for order effects – this is where participants are split equally between conditions, one group does condition A followed by condition B, with the second group doing condition B then condition A (remember, ABBA).
32
What is Matched-Pairs Design?
- Participants are paired together based on the same variables that can have an effect on the DV (e.g., participants may be matched on IQ for a memory study). They are then split across the conditions (one in condition A, one in condition B).
- This may be used in place of repeated measures designs – as participants in RMD may become aware of the true aims of the experiments if they experience all conditions of a study.
33
Evaluate Research Designs Independent groups design?
(+) **Order effects are not a problem** – Participants are also less likely to guess the true aims as they only contributing to one condition of the experiment.
(-) **Participants who take part in IGD’s are not the same** – If researchers find a difference between the groups on the DV, it could be due to differences between the participants rather than the effects of the IV. Random allocation may then be used to counter this issue.
(-) **IGD’s are less economical** – Compared to repeated measures – twice as many participants would be needed to produce the equivalent data that RMD’s produce as each participant only contributes a single result.
34
What are extraneous variables?
- Any variable other than the independent variable that can have an effect on the dependent variable if it’s not controlled for.
- They’re nuisance variables that **do not** vary systematically with the IV – they do not confound the findings of a study but can make it harder to detect a result.
- Where possible, researcher will attempt to identify them at the start of the research in order to take steps to minimalise their influence on the IV.
- Examples of EV’s include the age of participants, lighting in laboratories, educational background, time of day of testing, etc.…
35
What is an Aim?
A general statement of what the researcher intends to study. It is the purpose of the study and is based on the initial ideas that need exploring. It is stated at the start of the study.
36
What are confounding variables?
- Any variable other than the IV that may have affected the DV – confounding variables mean that researchers cannot be sure of the true source of the change to the dependent variable.
- Confounding variables **do vary systematically** with the independent variable.
- An example of CV’s include the personality of participants – This would have to be controlled for, where personality types are investigated prior to the study to make sure there is an even mix of personality types across all groups/ conditions
37
What is a Quasi Experiment?
The experimenter is not able to randomly allocate the participants because IV is a feature of the participants.
38
What is Temporal validity?
- When research findings successfully apply across time. E.g. changes to attitudes towards gender roles could affect the temporal validity of
research from the 1950’s.
39
What is Population validity?
- Can the findings be reasonably generalised from the research sample to the wider population?
- E.g., if a research study only involved men or
American citizens, then it can be said to lack
population validity.
- External validity can be assessed and
improved by replicating a study at different times and places – if replication produce similar results, then it can be said that the results have external validity.
40
What is Construct Validity?
- Whether a measure is successfully measuring the concept it intends (e.g., does a questionnaire directly measure IQ?).
41
What is Internal Validity?
The degree to which an observed effect was directly due to the change manipulation of the IV, rather than the result of other factors (e.g., confounding/ extraneous variables).
There are two types of internal validity:
- Construct
- Concurrent
42
What is Internal Reliability?
43
What is Internal reliability?
- Describes the internal consistency of
measurement, i.e., consistency within itself.
- This may include whether the different items/ questions on a questionnaire are measuring the same thing.
- On way to measure this is by the split-half
method – data is collected and split in half
randomly. The results are then compared to see
if the results take from each half are similar. If
the results are similar, them this would indication that the measure has internal reliability.
44
What is external reliability?
- This assessed reliability when different measures of the same thing are compared, i.e., does one measure match up against another measure.
- Discrepancies will therefore influence inter-observer reliability as results could change if one interviewer conducts an interview differently to another research.
- **Reliability issues** can be improved by standardising procedures – making sure that procedures are carried out in the same way. This can be done through methods such as implementing interviewer training and/ or practice through pilot studies
45
Evaluate Repeated measures design?
(+) **Participant variables are controlled,
and fewer participants are needed** :
because participants take part in all
conditions of the study.
(-) **Demand characteristics** – DCs tend to be a feature of RMD’s – this is because participants are more likely to figure out the true aims of the experiments when experiencing all conditions of the study.
(-) **Issue of order effects** – Taking part in more than one task could lead to boredom or fatigue, which could influence participants performance in the next condition. On the other hand, participants performance could increase due to them having practice in the conditions, especially in tasks which measure skill-performance. Order effects as a confounding variable.
(-) **Participants take part in more than one condition of the experiment** – This is perhaps the biggest issue of RMD’s – it means that the change in IV from condition A can still continue to have an influence in condition B (e.g., if someone take antihistamines in condition A,
drowsiness from them can still influence participants performance in condition B). Counterbalancing is used to combat this.
46
Evaluate Matched Pairs Design?
(+) **Order effects and demand characteristics are less of a problem** – This is because participants only take part in a single condition.
(-) **They’re time consuming and expensive** – This is due to the time it takes to match participants and pre-tests are often required in order to match participants. MPD’s are also therefore less economical.
(-) **Participants could never be matched precisely** – Despite MPD’s trying to reduce participant variables, this is still a problem. There will still be important differences in participants that could have an impact on the IV.
47
What are the evaluation of Random Sampling?
(+) **Free from researcher bias** – the researcher has no influence over who is picked for the sample, which prevents them from choosing people who they believe will support their hypothesis.
(-) **Difficult and time consuming** – it can be very difficult to create and obtain a list for the target population.
(-) **Samples can still be underrepresented** – It is more likely to produce a more representative sample than opportunity sampling (for instance) but is is still possible to have difference groups over/ underrepresented (such as there being a high percentage of white British participants, or high percentage of females).
(-) **Some participants may refuse to take part** – therefore the sample become more of a volunteer sample (an issue with all sampling techniques).
48
What is Systematic Sampling?
- Every nth person is selected from the population, e.g., every 7th person is selected from a university register.
- Names are organised onto the list to create a sampling frame – the interval of sampling can be decided randomly in order to reduce bias.
- The researcher makes their way through the sampling frame until a sample is obtained.
49
Evaluate Stratified Sampling?
(+) **Avoids researcher bias** – Participants that make up the numbers of the strata are randomly allocated, with the researcher having no influence.
(+) **Produced as representative sample** – its designed to accurately reflect the composition of the population. This means it is more possible
to generalise the findings of research.
(-) **It is not perfect** – The strata’s
identified cannot accurately reflect all the
ways that people are different, so it is not possible to have complete representation
of the target populations.
50
Evaluate opportunity sampling?
(+) **Convenient** – time and effort is saved, and the method is much less costly (both in terms of time and money) than other methods (such as random sampling).
(-) **Suffer from two forms of bias** – firstly, the sample is under representative of the target population as it is drawn from one very specific area (such as from one university building). Findings therefore cannot be generalised to target populations. Secondly, the researcher has complete control over who they approach to take part in their study, and so (for example) may avoid people they either don’t like the look of, or who don’t believe they will support their hypotheses.
51
Evaluate Self-Selecting/Volunteer Sampling?
(+) **Easy** – minimal effort is required from the researcher as the participants ‘come to you’.
Therefore, it is less time consuming than other sampling techniques.
(-) **Only certain people may come forward** – Participants are likely to be highly motivated and may potentially have extra time on their hands.
Participants may be more highly motivated to be more helpful or may need the money (if taking part in research which involves monetary incentives) – this results in volunteer bias.
52
Evaluate Sys
53
Evaluation of Systematic Sampling?
(+) **Avoids researcher bias** – Participants that make up the numbers of the strata are randomly allocated, with the researcher having no influence.
(+) **Produced as representative sample
** – its designed to accurately reflect the composition of the population. This means it is more possible to generalise the findings of research.
