research methods Flashcards
what are the four types of experiments?
Laboratory
Field
natural
quasi
what’s a laboratory experiment?
Give a strength and limitation
manipulated IV
Artificial setting
strength
- Strong causation (controls for other extraneous variables (EVs))
limitation
- high EV
what’s a field experiment?
Give a strength and limitation
- manipulated IV
- natural setting
strength - High EV
limitation - good causation (some control over extraneous variables, not all as natural setting)
what’s a natural experiment?
Give a strength and limitation
- naturally occurring IV
- natural setting
strength - high EV, more ethical/practical.
limitation - low causation (little control over extraneous variables)
what’s a Quasi experiment?
Give a strength and limitation
- Naturally occurring IV
- setting is not relevant
strengths- more ethicalpractical.
limitation - reduced causation - more EV’s to explain IV, however some control over setting if artificial.
what are the four observational techniques?
structures vs unstructured
participant vs non participant
controlled vs naturalist
covert vs overt.
what is a structured and unstructured observation?
give a strength and limitation of both.
structured = behavioural categories
strength - high inter rater reliability
limitation - can miss some relevant info.
unstructured = no beh categories.
strength - don’t miss relevant info.
limitation - low inter rater reliability
describe participant vs non participant
give a strength and limitation of both.
participant: observer acts like a ppt.
strength - deeper understanding of behaviour.
limitation - unethical and impractical.
non-participant: observer is separate to pps
strength - shallow understanding of beh
limitation - ethical and more practical.
describe controlled vs naturalistic observations
give one strength and limitation of both.
controlled: artificial controlled setting
strength - high control and strong test- retest.
limitation - low EV
naturalistic: natural uncontrolled setting
strength - high EV
limitation - low control and weak test-retest.
describe Covert vs overt observations.
give one strength and limitation of both.
covert: pps unaware they are being observed.
strength - no observer effects
limitation - deception
overt: pps know they are being observed
strength - no deception
limitation - observer effects
what are the two sampling techniques for observations?
time
event
describe and evaluate time sampling
beh = sampled in time intervals, e.g. every nth min for nth mins.
strengths:
- useful if observation is over a long period of time.
- data recording = accurate if there are lots of behaviours as your less likely to miss them during the time interval.
limitations
- not useful if a short observation as unlikely to capture beh
- not accurate data as will miss behs when looking away and not recording.
describe and evaluate event sampling
every beh observed is sampled throughout entire observation. (never look away)
strengths
- useful is observation is for short time
- accurate data as will not miss any behaviours as recording all the time.
limitations
- not useful if observation is for a long time.
- will miss beh’s if lots of beh’s in short time period.
what are the two self-report techniques?
interviews
questionnaires
evaluate Questionnaires
strengths
- easy to distribute to a large number quickly.
- less socially desirable if questionnaire is anonymous.
limitations
- may misunderstand Questions and cannot gain clarification
compare open and closed questions in questionnaires
open - (how?) deeper understanding but harder to analyse the data
closed - (do? are?) easier to compare but lack of deeper understanding
evaluate interviews
strengths
- can clarify misunderstandings of questions and stop acquiescence.
- able to observe beh as well as self-report (triangulate data)
limitations
- more social desirability as face to face with interviewer (wanna look good)
- interviewer effect: may change responses based on subtle non-verbal cues.
compare a structured vs unstructured interview using evaluation.
structured - set questions asked, pre-prepared, no other prompts can be given.
strength - high test-retest, can be easily repeated with several pps and results compared easily.
limitation - more formal, like reading out a questionnaire. Less likely to talk freely and openly, less data.
non-structured - no set questions, free flowing conversational style.
strength - less formal, like a convo, can build rapport. More likely to talk freely and openly, so more data.
limitation - low test-retest, all questions will be diff, unique to pp. cannot be repeated in same way, harder to compare against other pps.
correlations
look at flashcards for correlation pics before exam incase get asked to say what one is.
evaluate correlations
strength
- useful when experiment would be unethical or impractical
- may suggest the need to follow up with an experiment
limitation
- lack of causation. No control over variables, cannot say one co-variable causes the other.
what are the two subsections of content analysis?
coding analysis
thematic analysis
what is coding analysis?
Words are chosen from the 1st transcript (text/film)
They are then counted in the next transcripts (actual sample) in a tally chart = quantitative data.
what is thematic analysis
Themes are chosen from the 1st transcript
Quotes are then selected in the next transcripts that fit the theme = qualitative
summarise content analysis
- produce a transcript
- read through transcript looking for reoccurring words.
- They are then counted in the next transcripts (actual sample) in a tally chart = quantitative data.
- read through 1st transcript again and choose themes from the codes.
- find quotes to fit the themes.
- form a conclusion
evaluate content analysis
strengths
- easy way to collect data without needing to collect participants.
coding analysis
- more objective use of data
- easy to compare against other sources of info.
thematic analysis
- Deeper understanding of the context behind the themes.
limitations
- more time consuming as many films/interviews must be transcribed.
coding analysis
- Lack of detail on the context behind the words.
- Researcher bias in the choosing of the words to count.
thematic analysis
- difficult to analyse
- researcher bias in interpretation of themes
- subjective interpretations.
evaluate the use of case studies
strengths
- Detailed understanding of the person (idiographic)
- Can be used to prove/disprove nomothetic experiments
- Have high EV as tend to be studied in natural setting.
limitations
- Unable to generalize (as unique)
- time consuming
is a directional hypothesis one or two tailed?
one-tailed
(non-directional = two-tailed)
give the structure for a directional hypothesis
IV1 will be higher/ lower then IV2
give the structure for a non-directional hypothesis
there will be a difference between Iv1 and Iv2 in the DV
describe how to do random allocation and why it’s done
- Select and write appropriate names on paper
- Place into a hat and shake
- Pull out a name without looking and allocate to Condition A
- Reshake hat and pull out a name, Allocate to Condition B
- Repeat until the correct no of pps have been allocated.
Reduces investigator bias and Increases generalisability as avoids the likelihood of similar people in one condition.
describe how to counterbalance (ABBA) and why it’s done
- Split the pps in half.
- ½ do Condition A then B
- ½ do Condition B then A.
Removes order effects from
a repeated measures design
what is standardisation and why is it done?
all variables in procedure = same
(same instructions, procedure, timings)
Controls for extraneous
variables and Increases the test-retest reliability.
what is a single blind trial what what does it control for?
pps unaware of aim of exp
reduces demand characteristics.
what is a double blind trial and what does it control for?
pp and the researchers are unaware of the aim of the experiment
reduces demand characteristics of pps, reduces investigator bias in researchers.
describe how to do volunteer sampling and give strengths and limitations for it
- Put an advert out, with aim &
procedure. - Pps approach the researcher to take part.
strength - Large sample size: an
advert reaches large audience, already consented, no drop out.
limitation - Unrepresentative: extroverts, retired or students more likely to volunteer
describe how to do opportunity sampling and give a strength and limitation
Use the pps available at the time.
E.g. professor using their own students.
strength - Practically easier: Captive audience already there. No effort required.
limitation -Unrepresentative: all
similar pps, e.g. students of high IQ
describe how to do random sampling and evaluate
- Put all names on equal sized paper
- Place them into a hat, shake
- Pull out name
- Re-shake hat & pull out name
- Repeat until the appropriate sample size is reached.
strength - More representative:
equal chance of selecting a variety of pp characteristics. Lack of human choice in the sampling method.
limitation - Unrepresentative: still a chance that there could
be a sample bias, e.g., people of similar characteristics.
describe how to do systematic sampling and evaluate
- Gather a list of pps
- Choose every nth name off the list
E.g., every 5th person until have the
appropriate sample size.
strength - Representative: no
human choice in the system, likely to have more varied characteristics.
limitations - Still unrepresentative: can still have sample bias, depends on the list and the order names are in.
describe how to do stratified sampling and evaluate
- Calculate the proportion of
characteristic(s) in the target pop. - Work out the same proportion of
characteristics in the sample.
E.g., If 60% female and 40% male, in a sample size of 10 = 6 females & 4 males.
strength - Most representative: it has been calculated to be similar to the target population. However difficult to match every characteristic.
limitation - Practically harder/low sample size: need to
acquire data on characteristics, calculate sample proportion and gain consent for those that match.
what are the three experimental designs?
- repeated measures
- independent measures
- matched pairs
describe and evaluate repeated measures design
Same pps take part in Condition
A & B
strength - No individual differences: same pps compared to themselves. Lack of participant variable differences.
limitation - Order effects: boredom, fatigue or practice effects from doing study twice.
can be controlled by counterbalancing.
describe and evaluate independent measures design
Different pps take part in Condition A & B
strength - No order effects: different pps so only do the study once.
limitation - Individual differences: comparing different pps with dissimilar participant variables
Controlled by MP design
describe and evaluate matched pairs design
Different pps but matched in
Condition A & B for a similar
characteristics.
strength
- No order effects: different pps, only do study once.
- Fewer individual differences:
matched for similar characteristics, so less pp variable differences.
limitation - Still some individual differences: Different pps and cannot match for every single characteristics, so pp variable
differences still.
Controlled through RM design
what is a pilot study?
a small-scale trial run of an actual investigation to check for any issues, modify the design or procedure which saves time and money.
pps given a questionnaire after the study to check for any issues,
e.g. for an experiment: check the instructions made sense, check timings and correct materials
For observation: check coding + beg categories = correct and specific enough, check timings/ location issues.
for self-report- remove confusing/ inappropriate questions, check questionnaire works - answers can be recorded correctly
what are the three aims of a peer review?
- allocate research funding
- Validate the quality and relevance of the research (i.e., whether it is socially sensitive).
- Suggest amendments or improvements to the research paper or study itself
describe how to do a peer review
- Editor starts the process;
allocating research to a reviewer. - Reviewer reads and critiques in a 2-4 week deadline.
reviewer critiques:
- Validity of the science, any scientific errors, design or methodology issues.
- importance of the findings to the scientific community
- Originality of the work and any
inaccurate/missing references
(plagiarism)
what are the four decisions that can be paid from a peer review?
- accept
- accept but amendments
- reject but with resubmission
- outright reject.
evaluate peer reviews
strengths
- Ensures that only high-quality research is published.
- Checks that all work is original and not plagiarised.
- Anonymity of reviewer can ensure honest critique.
limitations
- anonymity of reviewer’s = can criticise a rival researcher. An ‘open review process’ avoids this.
- Editors tend to publish significant results more than non; to increase the credibility of their journal (file
drawer problem). - Slows down rate of change in science – ‘new’ theories going against current opinion are less likely to be passed by a reviewer
what are the two types of reliability?
- inter-rater/observer reliability
- test-retest reliability
what is inter-rater reliability
- 2 or more observers individually
record behaviour - They compare the scores using a
correlation. - If there is +0.8 correlation
between the scores, there is high
inter-rater reliability.
what is test-retest reliability?
- Pps take the same test twice, using the same procedure at different times or settings.
- If there is a +0.8 correlation
between the scores there is high
test-retest reliability
what is population validity?
Whether the results from a sample can be generalised to other types of people in the normal population.
what is face validity?
Whether a test looks like it measures what it intends to measure.
what is concurrent validity?
- Compare a new measure against an existing valid measure.
- If there is a +0.8 correlation between the scores, the new measure has high concurrent validity.
what are the four features of a science?
- empiricism (direct observation)
- control (IV causes DV)
- objectivity (results not effected by subjective option)
- replicability (repeat study to get same results)
what is a paradigm?
Shared set of assumptions about a subject matter & methods used
what is inductive reasoning?
- observation
- produce theory
- create hypothesis
- draw conclusions
- observe
what is deductive reasoning?
- existing theory
- create hypothesis
- collect data
- draw conclusion
- accept/ reject hypothesis
describe the sections of a report?
- Abstract - summary sentence from all sections - enables reader to ensure research is appropriate.
- introduction - background review of psych research. Explains how the research fit into research already in existence.
- method - materials, design, sample, procedure. enables accurate replication of study for test-retest.
- results - descriptive, inferential stats, graphs, tables. shows data analysis for reader to refer to and check accuracy of interpretation.
- discussion - explains results and how results fit with previous research, limitations, future research. Shows researches implications.
- references - ensures credit is given to authors of research referred to, avoid plagiarism, allow reader to find original source.
how do you reference a journal article?
Surname, initial. (year). title of journal, volume/issue number, page number
how do you reference a book reference?
surname, Initial. (year). book title. publishing location: publisher
what is primary vs secondary data
primary = directly gathered data, e.g. experiments or observations
secondary = indirectly gathered info from other sources
e.g. meta analysis, study of many other studies results and conclusions drawn.
evaluate primary and secondary data
primary data
strength
- specifically related to the hypothesis being tested.
limitation - time consuming to collect and expensive
secondary data
strength - cheap to use, easy to conduct.
limitation - data isn’t specific to the hypothesis being tested.
how to calculate median and when is it effective to use
middle value when put in order.
ordinal data, not as effected by outliers.
how to calculate mode and when appropriate to use
most frequent number.
nominal data, not effected by outliers
how to calculate standard deviation
how far the data is spread from the mean
most precise measure (uses all data set)
how to calculate total percentage
(score/total score) x 100
how to calculate percentage increase?
(new-original)/original) x 100
how to calculate percentage decrease?
((original-new)/original) x 100
when would you use a bar chart?
categories (IV)
experiments
when would you use a histogram?
continuous data
distribution of data
when would you use a scatter graph?
correlation
co-variables
when would you use a table?
represents descriptive stats
what is a normal distribution?
when the mean, median and mode are all equal.
what’s a positive skew?
the mode is lower than the median and mean.
what’s a negative skew?
the move is higher than the median and mean.
give reasons for a skewed distribution
measure is too easy (neg skew) or difficult (pos skew)
sample bias - opportunity or volunteer sample
outliers in data, affecting the mean
how to normally distribute a skewed distribution?
- make the measure harder (neg skew) or easier (pos skew)
- reduce the sample bias, e.g. stratified or random sample.
- remove outliers in data
describe interval data
Data has fixed intervals between each score. It is on a scale or has a fixed 0.
e.g. time, height, score, number of
describe ordinal data
Ranked based on subjective data, the intervals between each pp score varies.
e.g. rating of colour preference/ attractiveness
describe nominal data
data in categories or labels. Tends to be binary data
e.g. yes or no, agree or disagree, old or young.
how would you convert the age data in years (interval data) to ordinal and nominal data.
(interval) age data in years
(ordinal) rank age - 1 being youngest, 2 next oldest etc.
(nominal) age below 30 = young
age above 30 = old
how would you covert heart rate of racing drivers in bpm to ordinal and nominal data.
(interval) heart rate of racing drivers bpm
(ordinal ) rank bpm 1 being lowest, 2 next highest.
(nominal) below 50 bpm = slow, higher than 50 = fast.
practice drawing sign test table!
practice sign test maths questions!
critical value questions!
what is the level of significance in psychology? what does this mean
p<0.05 (5%)
95% confidence there is a significant finding. only 5% probability that the results are due to chance.
balance between probability of making a type 1 and type 2 error.
what is a type 1 error?
false positive
rejected the bulk when should have accepted it
more likely at 10% level of significance
what is a type 2 error?
false negative
accepted null when should have rejected it.
more likely at 1% level of significance
when would we use a 1% level of significance?
if the research is socially sensitive as it’s better to be cautious and make a type 2 error (could be damaging to society if incorrect)
practice 12 markers
