Investigation design

The Independent variable: The variable that the researcher manipulates and which we assume has a direct affect on the IV.
The dependent variable: The variable that is affected by changing the IV.
Aim: General statement about the inspected outcome of an investigation.
Hypothesis: A precise, testable statement of what you expect or predict.
1. Directional: Specifically states the direction of the results.
  1. Usually used when previous research findings suggest which way the results will go.
2. Non-directional: Direction of results is not specified and could go either way.
  1. When previous research findings are inconclusive.
3. Null hypothesis: It makes a very precise prediction (Nothing will happen).
Operationalised: Variables must be operationalised. This means describing the process by which a variable is measured. (How are you going to measure it).
A good Hypothesis is: PTM - Precise, testable and measurable.
Extraneous variables: A variable other than the IV that produces a change in the DV. If this happens then the results of the study are said to be confounded.
1. Situational Variables: Connected to the research situation e.g. temperature, noise, time, instructions.
  1. Controlled by standardised procedures and standardised instructions.
    1. Standardisation is where all conditions are kept the same in a study, to ensure that each participant has the same overall experience.
2. Participant variables: Connected with the research participant e.g. gender, age, IQ, familiarity with task.
  1. Controlled by repeated design (AGE), independent measures (GENDER), matched pairs measures (IQ) and random assigning to conditions.
3. Experimenter variables: related to the researcher e.g. gender, personality, apperance.
  1. Controlled by keeping the same researcher in all conditions.
4. BLT
  1. Counterbalance conditions by making one half to perform A first and then B and the other half to perform B first then A.
5. Distraction or confusion.
  1. Standardised instructions should be given in a clear and simple form.
Demand Characteristics: Participants form an idea about the purpose of the study which could have an effect on their responses and make the conclusions drawn from the study inaccurate. They could also be influences such as coffee, previous sleep etc.
1. Guessing purpose.
2. Annoy researcher by giving wrong answers.
3. Acting unnaturally due to nervousness or to look good.
4. Overcome demand characteristics by...
  1. Single blind procedure - the subject does not know which condition of the experiment they are in.
  2. Deception - To hide the aim of the research but will create ethical issues that will need to be overcome.
Investigator effects: Where the actions of a researcher affect the outcome of a study in an undesirable way.
1. Physical characteristics.
2. Accent, tone, non-verbal communication.
3. The investigator could be bias to the interpretation of the data.
4. Double blind procedure means neither the participant or the investigator knows which condition the participant is in.
Pilot Study: Aim is to check the method and to find solutions to any issues. A small scale trial run procedure to identify any flaws and areas for improvement before time and money are invested.
1. Experiments
  1. Instructions are clear.
  2. Dependent variable covers full range of scores to avoid floor and ceiling effects.
  3. Does it lead to demand characteristics.
2. Self reports (questionnaires/ intervies)
  1. Understandable questions.
  2. Closed questions offer suitable options.
  3. Are open questions needed to allow for unexpected responses.
3. Observations
  1. Operational (working) definitions of behavioural categories.
  2. Behavioural categories do not overlap.
  3. Are participants effected by the observer?
Reliability
1. Refers to the consistency of results.
2. Internal - How consistent a text is within its self.
  1. Split - Participants are split into two groups, if all score similarly on both halves, the questions measure the same thing. A high correlation between scores shows a high reliablility.
  2. Inter-observer reliability - when doing observations, need to make sure behavioural categories are understood. and should get a high correlation in observer results.
3. External - How consistent a test is over time and location.
  1. Test/retest - Test again at a different TIME or LOCATION. should get similar or same results for high external reliability.
Validity
1. Means we are measuring what we say we are measuring.
  1. Face validity - quick assessment to see if test 'appears' to test variable.
  2. Content validity - a closer look comparing with set standards of measuring the variable.
  3. Concurrent validity: Compare stores on test to an already established test - should get similar scores if new test is valid.
2. More valid results if extraneous variables are controlled.
3. Internal results: The test is measuring what we intend to measure.
4. External (Ecological): The findings can be applied to real life settings NOT just the research setting.
Experimental design
1. Repeated measures: Involves using the same participants in each condition of the experiment.
  1. No participant variables.
  2. More economical as fewer participants required.
  3. Order effects.
  4. Demand characteristics.
2. Independent group measures: Involves using different participants in each condition of the experiment.
  1. Order effects do not influence second condition.
    1. Boredom Tiredness Learning
  2. Demand characteristics are less of a problem.
  3. Participant variables.
  4. More subjects = less economical.
3. Matched pairs: Involves using different but similar subjects in each condition of an experiment.
  1. Reduced participant variables.
  2. Order effects do not occur.
  3. More subjects = less economical.
  4. Matching subjects is difficult and time consuming.
Sampling
1. Random: Truly random sampling only occurs when every member of a target population has an equal chance of being selected.
  1. Unbiased.
  2. Representative.
  3. Very rare as impractical.
  4. The larger the target population, the more difficult to sample.
2. Opportunity: Selecting those available at the time.
  1. Quick and convenient.
  2. Economical.
  3. Could be biased on the part of the researcher.
  4. Unrepresentative samples so cannot generalise.
3. Volunteer (self-selecting): Those who have consciously or subconsciously volunteered.
  1. Relatively convenient.
  2. Ethical.
  3. Subject could be biased
  4. Volunteers are unlike non-volunteers in many ways.
Ethical Issues (CCOWPD)
1. CONSENT: Fully understand what they are agreeing to.
  1. Debriefing, signature at start, under 16 and vulnerable people need consent from carer.
    1. Debriefing: A way of informing participants about the nature of the study. It is better to tae place after the study.
2. OBSERVATIONAL STUDIES: Avoid invading privacy buy observing them in places they would expect to be observed.
  1. Only observe people where they would expect to be observed.
3. WITHDRAW: To protect participants from harm, give them the right to withdraw from study at any point.
  1. Debriefing
4. PROTECTION: Protect participants from both physical and psychological harm.
  1. Debriefing and withdraw those in harm.
5. CONFIDENTIALITY: Unless agreed in advance, their results and personal info should be completely confidential.
  1. Debriefing to reassure.
6. DECEPTION: Participants must be told about the aim as soon as possible.
  1. Debriefing.

Next up

Investigation design

Description

Resource summary

Similar

	Created by emma.clissold about 11 years ago