research methods

  • Created by: f00233
  • Created on: 27-05-18 18:33

TYPES OF VARIABLES

Independent-

the thing that you manipulate in a study to see the effect it has on another variable 

Dependent-

the thing that you measure in a study

Operational-

the specific way in which its measured in the study

Extraneous-

any variable apart from the iv that may affect the dv, impacting the validity. anything that could impact the perfromance(situational; time, weather, noise. participant; age, mood, IQ)

Confounding-

have caused an influence on the study 

1 of 20

TYPES OF VALIDITY

external-the degree to which the findings can be generalised to other settings

internal-the degree to which an observed effect was due to manipulation or other variables 

ecological-the extent to which the results can be generalized to real life settings

face-when a study completes its aim

population-how well a sample can be generalized to the whole world/target population 

temporal-refers to the validity of the findings in relation to the progession of time (is it a "child of it's time"?)

2 of 20

TYPES OF EXPERIMENTAL DESIGN

  • Repeated measures, each participant receives all the conditions

- the order of the conditions may impact results=order effects 

- can guess the purpose=demand characteristics 

  • Independent groups, only participates in one condition

- can't control the effectsof participant variables 

- needs double the amount of participants= costly

  • Matched pairs , two groups matched on key characeristics, only participates in one condition 

- time consuming and difficult to match 

- not possible to control all participant variables 

  • you can... counter balance (ABBA), randomly allocate( number from hat) conduct pilot study
3 of 20

TYPES OF EXPERIMENTS

  • labatory 

conducted in a special environment with high control

normally artificial tasks, participants are aware they're in a study 

+ high control over variables, high internal validity, standardised  

- low ecological validity, demand charcteristics, can't be generalized  

  • field 

conducted in everyday setting of participants

IV is still manipulated, participants not usually aware 

+ higher ecological validity

- hard to control variables, ethical issues 

4 of 20

TYPES OF EXPERIMENTS

  • natural

conducted in everyday environment, but the IV is naturally occuring

+ high ecological validity, less demand characteristics, allows research on normally unethical things, get to study real life problems

-only a correlation as IV isnt manipulated, no control over variables, threat to internal validity 

  • quasi 

the IV is naturally occured but the DV is measured, the IV has not been made to vary but its pre existing features (gender, ethnicity, age, LoC, sexuality) 

+ allows comparisons between different types of people 

- demand characteristics, impacts internal validity, normally artificial task= low ecological validity

5 of 20

TYPES OF SAMPLING

  • opportunity those convenient or avaliable  

+ easy, less time consuming

- biased (confident ones or those who don't work) 

  • random random number generator, lottery method, random table number

+ unbiased, everyone has an equal chance of selection

- time consuming

  • volunteer advertise the study 

+ access to a variety, less biased and more representitive

- volunteer bias (more helpful, confident, more time)

6 of 20

TYPES OF SAMPLING

  • stratified

subgroups within a population are identified,selected through random allocation

+ representative 

- very time consuming

  • systematic

pretermined system (every nth person), the interval is applied continuously

+ unbiased

- bias unless you use random selection, then every nth person 

7 of 20

ETHICAL ISSUES

  • informed consent
    • RPOV, revealing true aims, but then they may guess (demand characterisitcs)
    • PPOV, basic human right of whether to participate 
  • deception
    • RPOV, necessary to hold truth, you can withhold some aims but not provide false aims
    • PPOV, unethical, leads to psychologosists seen as untrustworthy,wont want to participate
  • right to withdraw 
    • RPOV, if they leave then the strong ones are left, biased sample
    • PPOV, withdraw if uncomfortable, money or a reward are given which forces them to stay
  • confidentiality
    • RPOV, may be difficult if they're wanting to publish results, may only guarantee anonimity
    • PPOV, only acceptable for a persons data to be recorded if the data hides the participant
  • privacy
    • RPOV, hard to avoid if you're doing a field/natural study 
    • PPOV, they won't be expecting to be studied, invasion of privacy
  • protection from harm
    • RPOV, vital questions may cause distress, difficult to predict outcomes, not guaranteed
    • PPOV, must leave the study in the same state they arrived, no extra harm 
8 of 20

DEALING WITH THESE ISSUES

  • BPS guidelines
  • cost benefit analysis
  • ethics committee
  • punishment of psychologist
  • informed consent= presumptive, prior, parental consent
  • deception= debriefing
  • right to withdraw= told at start their right 
  • protection from harm= stop study if there's harm, avoid risks greater than everyday life
  • confidentiality= shouldn't record names, just as numbers
  • privacy= don't study anyone without informed consent 
9 of 20

PROBLEMS WITH EXPERIMENTS

  • demand characteristics
    • the totality of cues that convey the experimental hypothesis to the participant that becomes determinates of the participants behaviour- Orne
    • in experients, participants are unsure of what to do, they actively look for clues to see how they should behave. 
    • acts as an extraneous variable 
  • investigator effects
    • any cues from an investigator that encourage a certain behaviour which will lead to the fulfiment of the experimentors expectations
    • indirect- refers to situations where an investigator may not clearly specify the standardised procedure, leaves room for the results to be influenced by the experimenter.
  • dealing with these issues
    • single blind design
      • participant is not aware of the aims/condition they're recieving
    • double blind design
      • both participant and experimentor are blind to the aims
    • experimental realism
      • if the task is engaging they'll focus on the study, not that they're being observed 
10 of 20

OBSERVATIONAL TECHNIQUES

  • naturalistic
    • natural situation where everything is left as it was 
    • researcher does not interfere with the environment 
    • observing people in their natural environment (e.g child in nursery)

+ realistic, high ecological validity

- little control over other things that may impact their behaviour, unethical 

  • controlled
    • variables are regulated by the researcher 
    • allows them to investigate the effects of certain things on another

+ able to establish cause and effect, highly controlled

- lack validity, demand characteristics if it isn't natural 

11 of 20

OBSERVATIONAL TECHNIQUES

  • overt, participant knows they're being observed 

+ can gain informed consent 

- investigator effects, demand characteristics, social desirability bias

  • covert, participants unaware they're in a study

+ behaviour is natural, investigator effects are unlikely

- unethical

  • participant, researcher is part of the group being observed

+ obtain in depth data 

- presence might influence their behaviour, could be risky and dangerous is caught 

  • non-participant, observes from a distance and doesn't interact 

- might overlook or miss important behaviours

12 of 20

OBSERVATIONAL DESIGN

  • unstructured observation
    • record all relevant behaviour with no system 
    • too much to record, only record the eyecatching/visible behaviour 
    • may use this approach as a pilot study for a structured observation later on 
  • structured observation
    • aims to be objective and rigorous
    • main two ways of observing is behavioural categories or sampling procedures
  • behavioural categories
    • should be objective
    • cover all possible behaviours without having a waste/other category
    • be mutally exclusive/specific, the behaviour shoudl fit into only one category
  • sampling procedure 
    • event sampling
      • counting the number of times a certain behaviour occurs in a target individual
    • time sampling
      • recording behaviours in a given time frame 
13 of 20

SELF REPORT TECHNIQUES

  • questionnaire
    • set of written questions, to gain both quantitative and qualitative data, always structured

+ distributed to large number, high pop validity, reduces social desirability bias if anonomous

- biased sample, only those who can read or write or those with extra time 

  • structured interview
    • pre-determined questions, a face-to-face questionnaire 

+ standardised, easily repeated, 

- interviewer bias, social desirability bias 

  • unstructured interview 
    • new questions are developed during the interview, known also as a "clinical interview"

+ more detail obtained

- social desirability bias, interviewer effects, trained interviewer=more costly 

14 of 20

SELF REPORT DESIGN

  • questionnaire construction
    • writing good questions
      • clarity
      • bias 
      • analysis 
        • open/closed questions
    • writing a good questionnaire 
      • filler questions
      • sequence for the questions
      • no leading questions
      • pilot study 
  • design of interviews
    • recording the interview 
    • the effect of the interviewer 
      • non verbal communication
      • listening skills
    • questionning skills in unstructured interview 
15 of 20

CORRELATIONS

  • correlation
    • systematic assosiation between two continous variables
    • positive= two variables increase together 
    • negative=one variable increases as another decreases 
    • zero=no relashionship between the variables 
  • scattergrams
    • way of presenting a correlation
    • the scatter of the dots indicate the degree of correlation
  • correlation coefficient 
    • a measure of the extent of correlation that exists between the two variables 
    • it is a number, maximum being +1, lowest being -1
    • tells us how closely the variables are related 
    • to find out if it is significant, you have to use the table
16 of 20

MEASURES OF CENTRAL TENDENCY AND DISPERSION

  • central dendency, tells us the central value of a set of data 
    • mean
      • sensitive, takes account of all numbers= easily distorted 
    • median
      • not affected by extreme scores, not sensitive enough
    • mode
      • unaffected by extreme values, not useful when there are multiple modes 
  • measures of dispersion, tells us how spread out the data is 
    • range
      • easy to calculate, affected by extreme values, fails to take into account the distribution
    • standard deviation
      • precise measure of dispersion, easy to calculate, may hide some of the extreme values 
17 of 20

DISPLAY OF QUANTITATIVE DATA AND DATA DISTRIBUTION

  • displays 
    • tables
    • bar charts
    • histograms
    • line graph 
    • skattergram
  • data distribution
    • normal
      • classic bell shaped curve 
      • predicted distribution when considering an equal likely set of results
    • skewed
      • not equally distributed around the mean 
        • postive (right foot shaped)
        • negative (left foot shaped)
        • think the shape of a foot (big toe to little toe)
18 of 20

TYPES OF DATA

  • qualitative
    • how/what they feel and think
    • in depth detail, hard to analyse
  • quantitative
    • how many, how long, how much 
    • forced answer, easy to analyse
  • primary
    • first hand data gathered by researchers themselves
    • time consuming, expensive
  • secondary
    • data collected from someone else earlier
    • quick and easy, economical 
19 of 20

PEER REVIEW

  • peer review
    • assesment of scientific by others 
    • serves 3 main purposes:
      • allocation of research funding
      • publication of research in academic journals and books 
      • assessing the research rating of a universities department 
  • evaluation
    • hard to find an expert to analyse or review it 
      • poor research is passed on due to them not understanding, smith 1999
    • anonymity is used so reviewers can be honest and objective
      • however as it is a competitive environment, they may review them badly 
    • pulications tend to be positive and paid for, don't have true information
      • publicator bias
    • it results in theorys that follow existing theories being published, rather than contraversial ones
    • still published if deemed poor research, we are eposed to unvalid research
    • as the people reviewing th work are in the smae field, they may prevent it from being published on purpose as there is limited funding, and there's high competition. 
20 of 20

Comments

No comments have yet been made

Similar Psychology resources:

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