Research Methods

  • Created by: Kerry
  • Created on: 09-05-15 12:06
What are the 5 features of science?
Empirical methods, objectivity, replicability, theory construction and hypothesis testing
1 of 100
What are empirical methods?
Gain information through direct observation or experiment NOT reasoned argument/unfounded beliefs. Collect facts, direct testing = empirical evidence
2 of 100
What is objectivity?
Not affected by expectations of the researcher. To be objective - control conditions where research is conducted (lab). Lab experiment best as can establish cause and effect by varying IV
3 of 100
What is replicability?
Repeat observation/experiment to demonstrate its validity. Same outcome confirms, especially if made by different person. Need to use standardised procedures so replication can be acheived
4 of 100
What is theory construction?
Facts alone = meaningless, need to construct theories to explain facts and observations. Help us understand and predict natural phenomena
5 of 100
What is hypothesis testing?
Hypothesis testing allows theories to be modified. Tests the validity of a theory. needs to be falsifiable - i.e. proved wrong
6 of 100
How is scientific approach reductionist?
Complex phenomena reduced to simple variables in order to study causal relationships between them.
7 of 100
Is psychology a science?
Most psychologists generate models that can be falsified/conduct well controlled experiments to test them. Miller - psychologists who attempt to be scientists are dressing up, best as a pseudoscience
8 of 100
Are goals of science appropriate for psychology?
Psychological approaches to treating mental illness have had success - goals of science not always appropriate
9 of 100
How is scientific approach deterministic?
Searches for causal relationships - ignores certain factors
10 of 100
What are the two methods of acquiring knowledge in the scientific process?
Induction and deduction
11 of 100
What is induction?
Reasoning from the particular to the general. May observe natural phenomenon and come up with a theory. Example is Newtons Laws
12 of 100
What is the process for induction?
Observations --> testable hypothesis --> conduct study to test hypothesis --> draw conclusions --> propose theory
13 of 100
What is deduction?
Reasoning from general to the particular, start with theory and look for instances that confirm it. Darwins theory of evolution is an example
14 of 100
What is the process for deduction?
Observations --> propose theory --> testable hypothesis --> conduct study to test hypothesis --> draw conclusions
15 of 100
What did Popper argue?
Deductive method best to use - propose theory then seek evidence to support or contradict. Researchers can seek falsification - shows theory has been tested properly. If theory is hard to disprove - it is good
16 of 100
Why did Kuhn criticise Popper?
Argues he possessed idealized view of science, science progresses differently in real world. When theory is falsified - science still clings to it known as a paradigm. When enough evidence or alternative theores - paradigm shift, change and progressi
17 of 100
What is a peer review?
Assessment of scientific work by others who are experts in the same field - ensures that any research conducted and published is of high quality
18 of 100
What did The Parliamentary Office of Science and Technology suggest was the purpose of peer review?
Allocation of research funding, Publication of research in scientific journals and books, Assessing the research rating of university departments
19 of 100
How is it related to research funding?
Government/charities pay for research, public bodies require reviews to decide which research is most worthwhile
20 of 100
How is it related to publication of research?
journals offer chance to share results, peer review used as means of preventing faulty or incorrect data entering public domain.
21 of 100
How is it related to university departments?
All university science departments conduct research, this is assessed in terms of quality (Research Excellence Framework REF). Future funding depends on receiving good ratings from peer review
22 of 100
What is an issue of peer review and new research?
May reject new research just because it doesnt reflect the current theories and findings
23 of 100
How can they be susceptible to publication bias?
Journals prefer to publish positive results, results in bias in published research leading to misconception of true facts. This is the file drawer phenomenon
24 of 100
What is issue with experts?
Not always possible to find a appropriate expert, poor research may be passed because the reviewer didnt fully understand it
25 of 100
How might reviewer bias affect peer review?
May stop a piece from being published, if expert doesnt agree or if research doesnt come from a particular institution it may not be published
26 of 100
What is an issue relating to the internet?
Sheer volume of information on internet, means new solutions are needed to deal with information on internet and maintain quality e.g. Wikipedia
27 of 100
When would you choose to use an interview/survey?
When you want to get detailed information or in depth responses, want a lot of peoples opinions
28 of 100
Problems with interviews/surveys?
Honesty/social desirability bias, hard to analyse responses - content/thematic analysis. Interviews can be time consuming
29 of 100
When would you choose observations?
Want to gather data without largely affecting participants behaviour OR when you are interested in how a behaviour may appear in a natural environment, provide rich picture of what people do
30 of 100
Issues with observations?
Demand characteristics, a lot of information to collect - need to decide which is most useful e.g. categories such as in strange situation. Observer bias - subjective analysis
31 of 100
When would you choose a case study?
Have a unique single person or small group of people you want to study, usually longitudinal
32 of 100
Issues with a case study?
Difficult to generalise any findings to wider population, may need to use past events which could be unreliable
33 of 100
When would you choose a correlational study?
When you are looking for a relationship between two variables, can use large data sets and be easily replicated
34 of 100
When would you choose a lab experiment?
Looking for cause and effect between two specific variables, controlled environment
35 of 100
Evaluation of lab experiment?
Low in external validity - less like everyday life, reduced internal validity - experimenter bias and demand characteristics, is high in replicability though as standardised procedures used
36 of 100
When would you use a natural experiment?
When you have a naturally occurring IV, cannot draw causal conclusions as cant manipulate IV, reduce validity as cannot randomly allocate p's to groups
37 of 100
When would you use a field experiment?
Want to study behaviour in more natural environment but still want control - experimenter effects reduced, demand characteristics may still be an issue
38 of 100
Opportunity sample and issues of?
Select p's by using those most easily available - easiest method, biased as only drawn from small part of target population, relatively quick compared to others
39 of 100
Volunteer sample and issues of?
Participants selected by asking for volunteers, place advert etc. Access variety of p's, biased as highly motivated/free time.
40 of 100
Random sample and issues of?
p's selected using a random number technique, all members of target population identified and then selected by lottery method of number generated. all have EQUAL CHANCE. - potentially unbiased, more time consuming
41 of 100
Stratified and quota samples and issues of?
Strata (sub groups) in population are identified, predetermined number taken from each in proportion to representation in target population. Stratified = random, quota = opportunity - more representative, possibly biased
42 of 100
Systematic sampling and issues of?
select every nth person, list of p's in target population first. unbiased but not truly random unless you start at a random person and select a random nth number
43 of 100
What can reliability be divided into?
Internal and external
44 of 100
What is internal reliability?
Consistency within a test - how consistent it is at measuring something
45 of 100
How is internal reliability assessed?
Using split half method - compare two halves of a test, if the test is consistent/reliable then each half should give same result. calculate correlation coefficient to compare two scores
46 of 100
How is internal reliability improved?
Select the test items that provide the greatest similarity/consistency, remove items and see if remaining give more positive correlation
47 of 100
What is external reliability?
The ability to replicate the test or study and find similar or the same (consistent) results each time
48 of 100
How can it be assessed?
Using test-retest method - give same test on two separate occasions using same/similar p's. long enough to forget answers but not so long they change opinion. In an interview the same interviewer must be used
49 of 100
How can it be improved?
Provide interviewer with better training, test questions could be ambiguous so p's dont give same answers - refine questions
50 of 100
What is internal validity?(experiment)
The controlling of variables within the study - is result due to manipulation of IV or EV's?
51 of 100
How can internal validity be improved?(experiment)
PILOT STUDY, Single blind design - p doesnt know aims of study/involved in study, double blind study - investigator cannot give cues (elimates DC's)
52 of 100
What is internal validity influenced by? (experiment)
Demand characteristics, experimenter bias (effects of expectations), investigator effects (giving cues), time of day, temp, noise, order effects, (situational variables), age, intelligence, motivation, experience, gender (participant variables)
53 of 100
What is external validity?(experiment)
The extent that the results can be generalised to other settings
54 of 100
How can external validity be improved?(experiment)
replicate with different groups of people in different settings, set p's in more natural setting
55 of 100
What is external validity influenced by?(experiment)
Mundane realism (how realistic task is to real life) and ecological validity (can be generalise to different setting)
56 of 100
What affects validity in observations/surveys?
social desirability bias, interviewer bias, leading questions and content validity
57 of 100
How can it be assessed?
Lie scale - ask questions that test truthfulness, if they lie on high proportion they may not be truthful on actual q's. concurrent validity - compare to valid test to see if similar results, predictive validity - see if the test actual is successful
58 of 100
How can it be improved?
Change questions
59 of 100
What ethical issues for humans according to?
BPS -British Psychological Organisation
60 of 100
What are the principles/issues?
Respect (informed consent, confidentiality, deception, right to withdraw), competence, responsibility (protection from harm, debriefing), integrity
61 of 100
What are ethical issues for animals according to?
Animals (scientific procedures) Act - animal research only takes place in licensed labs with licensed researchers
62 of 100
What are the conditions licenses are granted in?
results justify means (use of animals), research cannot be done using non-animal methods, minimum number of animals will be used,discomfort or suffering is kept to a minimum by anesthetics or painkillers
63 of 100
Why use animals?
Fascinating to study/findings may benefit animals, greater control and objectivity in research, cant use humans, have enough physiology and evolutionary past to draw conclusions and generalise them to humans
64 of 100
Issues around using animals?
Sentient beings - dont know if they experience pain or emotions, respond to pain but may not be consciously aware of it, speciesm - Singer - discrimination on basis of species no different to racial or gender so using animals is speciesm
65 of 100
What is a issue relating to animal rights?
Regan - no circumstances where animal research is acceptable, animals have right to be respected and should never be used.
66 of 100
How can this be challenged?
Having rights is dependent on having responsibilities in society - as citizens, animals have no responsibilities so have no rights
67 of 100
What are the 3 R's? (Russell and Birch)
Reduction (use fewer animals), Replacement (use alternative methods where possible), Refinement (use improved techniques to improve stress)
68 of 100
When should a bar chart be used?
Useful to show differences in data e.g. means, between groups. Bars are separate as data is DISCONTINUOUS
69 of 100
When should a histogram be used?
distribution of whole group of data, bars are together as data is CONTINUOUS, column area of bars represent the frequency of the score
70 of 100
When should a scattergram be used?
Relationship between two variables, CORRELATIONAL data, strength and direction of the correlation
71 of 100
What is probability?
Likelihood of something happening, expressed between 0-1 (0= event wont happen, 1=definetly will)
72 of 100
What are stats tests used for?
Work out probability of results being due to chance
73 of 100
Why is level of probability set at 5%?
In order to avoid accepting or rejecting our experimental hypothesis incorrectly, 5% of results due to chance. If we are too strict or lenient with data a type 1 or type 2 error can occur
74 of 100
What is type 1?
too LENIENT, higher level of probability than we should be looking at (10%), accept hypothesis when we should reject it
75 of 100
What is type 2?
too STRICT, reject hypothesis when we should accept it
76 of 100
What is a level of measurement?
The type of data - nominal or ordinal
77 of 100
What is nominal data?
Categorical data e.g. type of car etc
78 of 100
What is ordinal data?
quantities that have a natural ordering e.g. choice on rating scale, ranking of favourite sports (non equal intervals)
79 of 100
When is Chi squared used?
Nominal data and indepedent groups (test of difference) or test of association
80 of 100
When is Mann Whitney test used?
Ordinal data and test of difference (independent groups)
81 of 100
When is Wilcoxon Signed Rank Test used?
Ordinal data and test of difference (repeated measures)
82 of 100
When is Spearmans Rho test used?
Ordinal data and test of association
83 of 100
When is null hypothesis (there is no difference between this and this) accepted?
If observed value is less than critical value
84 of 100
What is the degree of freedom?
In all but Chi sqaured it is the number of p's. Chi = (rows-1)x(columns-1)
85 of 100
How do you know if it is two tailed or one tailed test?
Directional hypothesis = one tailed, non directional hypothesis = two tailed
86 of 100
What is a directional hypothesis and a non-directional hypothesis?
Non directional predicts difference but doesnt specify the difference, directional does specify difference
87 of 100
Difference between qualitative and quantitative data?
Qualitative = data in non-numerical form, can be summarised but not counted. Quantitative = data in numerical form, behaviour measured in numbers or quantities
88 of 100
Why is qualitiative data difficult to summarise?
Can use measures of central tendency or dispersion or graphs. Can only identify repeated themes
89 of 100
Why is qualitative data inductive?
Themes emerge that are based in data which may lead to new theories. Deductive is less common - researcher starts with pre-set categories or themes, see if findings are consistent with previous theories
90 of 100
Another issue with qualitative data?
Very lengthy analysis because it is iterative
91 of 100
First principle of thematic analysis?
Read and re-read data transcript dispassionately, try to understand meaning and perspective of p's
92 of 100
2nd principle?
Break data into meaningful units, small pieces of text that convey meaning
93 of 100
3rd principle?
Assign code to each unit - these are initial categories, each unit can have more than one code
94 of 100
4th principle?
Combine codes into larger categories or themes
95 of 100
After this?
Final report to discuss the emergent themes, draw conclusions which may include new theories
96 of 100
Issues of quantitative data?
Oversimplifies human behaviour, qualitative gives rich detail but more subjective
97 of 100
Writing up method - what to include for design?
Lab/field/natural experiement, independent groups, Rm/ MP's, IV and DV, extraneous variables controlled?, dealing with ethical issues
98 of 100
Writing up method - what to include for participants?
sample size, sampling methods, control groups
99 of 100
Writing up method = materials?
questionnaire, paper, pens, stopwatch, consent form, standardised instructions etc
100 of 100

Other cards in this set

Card 2


What are empirical methods?


Gain information through direct observation or experiment NOT reasoned argument/unfounded beliefs. Collect facts, direct testing = empirical evidence

Card 3


What is objectivity?


Preview of the front of card 3

Card 4


What is replicability?


Preview of the front of card 4

Card 5


What is theory construction?


Preview of the front of card 5
View more cards


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Research methods and techniques resources »