Research Methods Flashcards
Independent Variable
This is the variable that the researcher manipulates in order to determine its effect on the dependent variable
Dependent Variable
Variable that is measured
Extraneous Variables
Variable that COULD affect the DV
Confounding Variables
Variables that HAVE affected the DV
Operationalisation
IV and DV defined and stated how it should be measured
Lab Experiment
Completed in a lab
Field Experiment
Carried out in the real world
Natural Experiment
Naturally occurring IV - comparing behaviour in single-sex school compared to mixed
Quasi Experiment
Naturally occurring IV taking place in a lab - difference between people that already exists such as levels of testosterone
Non-Participant Observation
Researcher does not get directly involved with interactions of the participants
Participant Observation
Researcher is directly involved - also works out whilst collecting information at a gym
Covert Observation
Psychologist goes undercover and does not reveal their true identity - group does not know they are being observed
Overt Observation
Psychologist reveals true identity and might also state they are observing the group
Naturalistic Observation
Researcher observes participants in their own natural environment and there is no deliberate manipulation of the IV
Controlled Observation
Researcher observes participants in a controlled environment and this allows for manipulation of the IV
Structured Interviews
Standardised order
Quantitative data
Closed questions
Unstructured Interviews
Informal in-depth conversation
Unplanned questions
Qualitative data
Open questions
Questionnaires
Closed/open questions
Must avoid:
Ambiguous questions and answers
Leading questions
Simple questions
Correlation Coefficient
Strength of a correlation, falls between -1 and 1
Correlation
Positive or Negative
Shown in a scattergraph
Null hypothesis
IV will have no effect on the DV
Non-directional/two-tailed
Does not state what the effect of the IV is on the DV but suggests there is an effect
Directional/one-tailed
Predicts the effect of the IV on the DV
Alternative/experimental
Predicts that IV will have an effect on the DV
Random Sampling
Every member of a target population has the same chance of being selected through RNG
Systematic Sampling
Participants are selected by taking every Nth person from a list
Stratified Sampling
Involves classifying the population into categories and then randomly choosing a sample which consists of participants in the same proportions as they are in the real world
Categories - age, gender
Opportunity Sampling
Involves selecting participants who are readily available and want to take part
Volunteer Sampling
Involves people volunteering to take part in a study through the researcher advertising their study via leaflets, posters, radio
Pilot Studies
Small scale investigation of the real study in order to identify any flaws in the structure
Independent Groups Design
Different people used in each condition
Randomly allocated participants to each condition
Repeated Measures Design
Each participant is tested in all conditions
Matched Pairs Design
People with matched characteristics that are important to the study are put into opposite groups
External Reliability
Whether a test results are consistent over time
Test-retest method can be used to check
Test-Retest
Test done again to see if results are similar
If they are, the results are reliable
Internal Reliability
Whether a test and the results are consistent within itself
Split-half technique used to check
Split-Half
The questionnaire is split in half and if participants score similarly on both halves of the questionnaire then the questions are measuring the same factors and there is internal reliability
External Validity
Ecological, temporal, participant validity
Internal Validity
When the outcome of the study is a direct result of the manipulation of the IV and there are no confounding variables
Presumptive Consent
Consent gained from people of a similar background to the participants in a study
Prior General Consent
Involves participants agreeing to be deceived without knowing how or when it will occur
- for example, “Would you be willing to take part in a future study based on memory, whereby the true aim of the study might be withheld?” If the participants agree, then you can conduct the study knowing that you have gained their informed consent prior (or before) the actual study
Retrospective Consent
Asking participants for consent after they have already participated in the study
If they don’t consent - data is destroyed
Protection from Harm
Investigators have responsibility to protect participants from physical and psychological harm during the study
Psychologist must stop any study immediately if they suspect a participant may be harmed
Right to Withdraw
Participants can leave study at any time
Participants can withdraw their data at any point in the future
Confidentiality
Participants’ data is confidential and should not be disclosed to anyone
Numbers or letters should be used instead of names
Deception
Ethics committee will approve or disapprove of experimental methods involving deception using a cost-benefit analysis test
Nominal (Discrete) Data
Data are in separate categories such as grouping people according to their favourite TV show or eye colour
Ordinal Data (Continuous Data)
Data that is ordered in some way
Listing order of favourite music genres
Interval Data (Continuous Data)
Data is measured using units of equal intervals such as miles or centimetres
Primary Data
Information observed or collected directly from first-hand experience
Primary data provides the exact type of data the researcher is looking for
Secondary Data
Information that was collected for another purpose
Researcher could use the data collected for another study
Usually less reliable than primary
Meta-Analysis
Combining results from a number of studies on a particular topic to provide an overall view
Mean
Most accurate measure and it takes into account all the scores
Can be distorted by a single extreme value
Mean score may not even be a value obtained
Median
Unaffected by extreme scores
Not as sensitive as the mean because not all scores are used in the calculation so it can be unrepresentative of the data if scores are clustered around high/low levels
Range
Quick and easy to calculate
Easily distorted by extreme values
Standard Deviation
Takes into account all the scores
More difficult to calculate than the range and can only be used on interval data
Inferential Statistics
When you assume that the results obtained from one study on one category of people will be the same for another category
So if Coke improves memory for women it will also improve memory for men
Descriptive Statistics
Measures of central tendency and dispersion
Level of Statistical Significance
Level at which the decision is made to reject the null hypothesis
Certain that the IV is having effect on DV and is not due to chance
Chance
Something has no real cause, it just happens - no real cause
Level of significance
Chance of results being due to chance/fluke is p<0.05 (5%)
Used when there is a directional one-tailed hypothesis
p<0.01 when research findings are critical and can be difference between life or death
Sign Test
State hypothesis
Count number of + and -
Choose less frequent sign
Identify N (don’t count participants with 0)
Select critical value
Compare S value to critical value
If S is equal to or less than critical value the results are significant and we accept experimental hypothesis
Report - Title
Should provide clear focus of the study and should involve the key variables that are being investigated, should not be too vague or too specific
Report - Abstract (150-200 words long)
Provides clear and concise summary of the entire investigation
Includes information such as background research, aims, hypothesis, methodology, experimental design, sample used
Type 1 Error
Occurs if an investigator rejects a null hypothesis that is actually true in the population
Type 2 Error
Occurs if the investigator fails to reject a null hypothesis that is actually false in the population
Correlation Advantages
- This technique does allow psychologists to establish the strength of the relationship between two variables and measure it precisely.
- This technique also allows researchers to investigate things that could not be manipulated experimentally for ethical or practical reasons.
- Once a correlation has been conducted predictions can be made about one of the variables based on what is known about the other variable.
Correlation Disadvantages
- Correlational analysis cannot demonstrate cause and effect; we cannot tell which variable influences the other.
- Even if there is a correlation between two variables it may be the case that the variables are not actually related but that there is a third unknown variable which influences both (confounding variable)
- Correlations can only measure linear relationships and does not detect curvilinear relationships. This is when there is a positive relationship up to a certain point but after that the relationship becomes negative or vice versa.
Replicability
The extent to which the findings of research can be repeated in different contexts and circumstances
Purposes of Replicability
a) Guarding against scientific fraud
b) Researchers can check to see if results gained were “a one off fluke” possibly caused by extraneous/confounding variables
c) If research findings can be repeated, we would say that the findings are reliable
d) Replicability can also indicate that research findings are valid
Falsifiability
The notion that scientific theories can potentially be disproved by evidence, it is the hallmark of science. It refers to proving a hypothesis wrong
Nominal and Independent
Chi squared
Nominal and Repeated
Sign
Nominal and Association
Chi squared
Ordinal and Independent
Mann Whitney - Less than or equal to be significant
Ordinal and Repeated
Wilcoxon
Ordinal and Association
Spearman’s Rho
Interval and Independent
Unrelated t-test
Interval and Repeated
Related t-test
Interval and Association
Pearson’s R
