Research Methods
- Created by: marmellow22
- Created on: 07-05-15 18:29
Planning Research
- Formulate research question
- Formulate aim- based on theory and should allow predictions
- Formulate hypotheses
- Needs to be operationalise the variables
Hypotheses
- Research (H1)= Directional~ will increase/decrease
- Null (Ho)= No difference
- Alternative (Ha)= Non directional~ will be a significant difference Used when primary hypothesis is the Null
Variables
- Independent (IV)= Cause (object changed/manipulated)
- Dependent (DV)= Effect (object measured/observed)
- Confounding (CV)= Variable that may have effected the DV, that is not the IV
- Extraneous (EV)= Variable that could affect DV that is not the IV
Ways of controlling Extraneous Variables
Standardisation:
- Keepimg variables the same e.g. participants, environment, tasks, measures
Counterbalancing:
- Ensuring variables occur in all possible combinations an equal number of times.
- Danger of Order Effects- Practise Effect (improve), Fatigue Effect (worsens) and Demand Characteristics (behave in the way they believe they are expected)
Randomisation:
- Deciding the order or use of variables by chance
Types of Experiment
Laboratory
- Carried out in a controlled environment
- Strength: Allows random allocation (each participant has an equal chance of being in each condition) when appropriate to ensure findings are not biased
- Limitation: Findings tend to lack Ecological Validity
Field
- Carried out in a natural environment
- Strength: More Ecological Validity
- Limitation: No control over environmental factors making it more difficult to reliably establish cause and effect
Quasi
- Experimentor does not directly control the IV when it is naturally occuring e.g. the Weather
- Can take place in a laboratory or in the field
- Less ethical issues as participants are not being manipulated as much as in Laboratory or Field
Experimental Designs
Experiments will have at least two conditions;
- Experimental: Condition where variable is tested
- Control: Condition that acts as a comparison; nothing changes
After establishing conditions need to choose an Experimental Design: way participants are used in conditions within an experiment
Types of Experimental Design
Repeated (related) Measures
- Same participants are used in all conditions
- Strength: Differences between conditions are likely due to the IV and not due to participant variables (differences between characteristics of participants
- Limitation: Order effects (reduced by counterbalancing or randomising condition order)
Types of Experimental Design
Independent Group
- Different participants randomly allocated to different conditions
- Strength: no order effects
- Limitation: Differences between may be due to participant variables. Larger sample (participants selected from the population) lower probability of a significant participant variables
Matched Pairs Design
- Different participants are used in each condition but are matched in terms of Key Characteristics
- Strength: No order effects
- Limitation: Time consuming and expensive (random allocation tends to give a balance of participants anyway)
Non-experimental methods
- Methods that don't involve direct manipulation of a situation or behaviour
- Used as they are seen as more Ethical
Self-Report Methods
Questionnaires
- List of pre-determined questions
- Strength: Easily made anonymous. More honest/valid responses given
- Limitation: Respondents may misunderstand/ misinterpret questions~ invalid responses
Self-report methods
Interviews
- Directly asking participants questions and recording their responses, often done on a one-to-one basis
- Structured: Pre-determined questions. Easier to compare interviewees' answers
- Unstructured: Start with some common questions and tend to be directed by participants. Interviewer can follow lines of enquiry and interviewees may introduce relevant ideas
Open and Closed Questions
Closed
- Participants offered fixed set of responses- Yes/NO response; Rating Scales (1-5); Multiple Choice
- Quantatative Data
- Easier to compare answers, helps researchers stay focused
Self-report methods continued...
Open
- Doesn't restrict responses
- Qualitative Data
- Better exploring answers in more depth- closer to truth
Pilot Studies
- Small scale, trial study to identify any factors that might negatively affect the outcome of a study
- Ceiling Effects- too easy
- Floor Effects- too difficult
Observational Studues
- Participants' behaviour recorded through watching them
- Carried out in a number of ways;
- Scoring System (1-5)
- Check list of Criteria
- Keeping a tally
- Making notes
- Video recording
Need to decide how observation will be carried out;
- Laboratory or natural setting
- Covert or Overt
- Participant or Non-participant
Laboratory
- Strength: High level of control- easier to establish cause and effect
- Limitation: Artificial environment means findings lack ecological validity
Observational Studues Continued...
Natural
- Strength: High level of Ecological Validity so findings should be generalisable to real life
- Limitation:Many uncontrolled variables making it difficult to draw any conclusion about causation
Covert
- Undercover observation
- Strength: Behaviour should be normal giving valid results
- Limitation: Ethical Concerns- is it right to watch someone without concent?
Overt
- Open observation (aware)
- Strength: Ethically sound but may withdraw themselves as a consequence
- Limitation: Observer effect- behave different making unreliable results
Observational Studues Furthered
Participant
- Joining in with activity
- Strength: allows researcher to experience a situation giving greater insight increasing validity
- Limitation: Researcher may become too involved and lose their ability to be objective
Non-partcipant
- Not joining in
- Strength: Researcher remains objective
- Limitation: researcher may not have a true understanding of behaviour being too removed from the situation
Correlation Studies
- Relationship shown graphically using a scattergram and measured by correlation co-efficient
Positive /
- Relationship between variables- both increase and both decrease
- Co-efficient of +1
Negative \
- Relationship between variables- one increases as the other decreases and vise versa
- Co-efficient of -1
Zero
- No relationship between variables
- Co-efficient of 0
Difference between correlation studies and Experiments: Correlation Studies cannot reliably establish cause and effect
Content Analysis- Psychology and the Media
- Interpreting secondary material as a means of understanding people
- Data can be quantitative (counting- number/percentage) or qualitative (more descriptive)
- Strength: few ethical issues- no or little direct contact with 'participants'
- Limitation: Easier to misinterpret their thoughts and behaviour
Qualitative Data
- Methods: Self Report (interviews); Describing Observations; Case Study
- Strength: Generally seen to have more construct validity
Quantitative Data
- Methods: Experiments; Structured Observations; Correlation
- Strength: Easily summarised into graphs or statistics
Sampling
- Sample of target population- wider group of people that research findings should apply to
- Sample is then projected onto the rest of the population
- Need to ensure there is not a biased sample
- Taken from the sampling frame that is within the Target Population
Types of Sampling Techniques
Random Sampling
- Every person in the sampling frame has an equal chance of being selected
- Strength: Avoids bias, once researcher has decided the nth Term, there is no control over who is selected
- Limitation: Chance (although unlikely) of a 'freak' sample
Types of Sampling techniques
Systematic Sampling
- Choosing every nth person in the sampling frame (every person in the sampling frame doesn't have an equal chance of being selected
- Strength: avoids bias, once researcher has decided the nth term, as there's no control over who is selected
- Limitation: Freak sample
Stratified Sampling
- Divided the sampling frame into key groups and selecting from each of these groups
- Strength: Avoids 'freak' sampling
- Limitation: Time-Consuming
Opportunity Sampling
- Participants chosen because they are conveinient
- Strength: Less time-consuming
- Limitation: Likely for biased as only certain types of people would volunteer
Ethics
Respect- for all individuals
- No more risk of harm than everyday life
- Negative effects of an investigation and long term consequences
- Debrief
- Use the smallest number of people as possible
Unethical Research is possible if psychologist can show that means justify the ends
Representing Data
- Tables graphs etc are immediate representation of data that is easy to read
- Bar Graphy (gap between)
- Line Graph
- Histograms (no gap)
- Scatter grams
Descriptive Data
Central Tendency
- Mode- most common
- Median- Ordered list and find the middle (2 numbers divided by 2)
- Mean- Add up and divided by number of figures
Measures of Dispersion
Range- Highest take away from the smallest
Standard Deviation
(Equation)
Related discussions on The Student Room
- AQA Sociology Unit 2 Education and Research Methods help »
- Psya4 research methods?? »
- Help with Psychology AS AQA A »
- Alevel physics endorsed practicals »
- How to revise research methods? »
- Oxbridge Research proposal for DPhil/PhD in History »
- AQA Psychology Revision help »
- Sociology AS level »
- How do I structure questions for SY3 and SY4 A2 Sociology WJEC »
- Crying for help.. »
Comments
No comments have yet been made