research methods mocks june
primary data
Refers To data that is reported by the original researcher; Either obtained from questionnaire interviews or witnessed firsthand
s: High level of control, Data generatered will fit the aim of investigation directly, increases for validity
l: Time and effort
secondary data
Information that was collected by other researchers, Already exists and has been subjected to statistical testing
s: Less time-consuming as data is already in the public domain
l: Concern over accuracy, Variability in the quality of the data might mean that it is bias
how consent analysis can be used to analyse a scenario?
Helps to summarise and classify communications in a systematic way
Researcher examines data and artefacts And familiarise themselves
Researchers identify coding units and they must be operationalised
Data is analysed by applying the coding units
A tally is made
how can you assess the reliability of content analysis?
Test retest reliability: Recoding the same data by the same researcher overtime to assess consistency
Interobserver reliability: Comparing coding Results from two or more observers to check the results are trustworthy
Pilot study could be used to analyse smaller artefacts
structured interviews
Predetermined set of questions, asked in same order for each interviewee (Standardisation)
Easier to analyse statistically due to quantitative data
Doesn't allow for additional questions to be asked
Unstructured interviews
More like a conversation where the interviewer is facilitating the discussion
Not asking set questions
Generates a large amount of qualitative data
more difficult to analyse statistically
semi structured interviews
Mostly prepared questions but allows deviation
Generates a large amount of qualitative data
More difficult to analyse
quantitative data
Numerical data that can be Statistically analysed
s: easy to analyse due to limited number of responses
More reliable, objective and less bias Due to established mathematical procedures
l: Lacks representativeness as it lacks meaning and contextual information, Lacks external validity
qualitative data
Non-numerical language based and expressed in words
Gives a unique insight into the human experience
s: Rich in detail and has enhanced external validity
l: Subjective as it relies on interpretations of the data, May lead to bias
ethical issues
Informed consent: Aims and consequences of research should be made clear before they agreed to participants
Deception: When information is deliberately withheld from Participants
Right to withdraw: Remove themselves all the data from a study at any stage
Protection from harm: Should Leave the research in the same state that they entered in
Privacy and confidentiality: The right to decide how information about them will be communicated to others, Where Participant personal information is protected by law
Debriefing: Should be fully debriefed until the true aim of the research
how informed consent can be obtained
participants should be aware of their right to withdraw and consequences of the research should be made clear before they agree
Presumptuive consent: taking a random sample, introducing them to the research (Consent is presumed)
Prior general consent: Agreeing to take part beforehand in numerous psychological investigations
Retrospective consent: Participants giving consent for their participation after taking part
Laboratory experiment
Conducted under controlled conditions
Researcher manipulates the IV to measure the effect on the DV
Participants are aware
High degree of control over extraneous variables, Increased internal validity, Cause and effect relationship can be established
Lack of external (ecological) Validity findings can't be generalised to settings beyond the laboratory
Field experiments
Naturalistic conditions
Researcher manipulates IV to measure the effect on the DV
Typically don't know they're taking part
High level of ecological validity, Greater mundane realism, Results are more likely to be representative of every day life
Less control over extraneous variables due to the natural setting, Cause an effect relationship is difficult to determine
natural experiments
Does not manipulate the IV but examine the effect of an existing IV on a DV
Higher level of external (ecological) validity, Demand characteristics do not impact the experiment
No control over the environment and extraneous variables, Difficult to accurately assess the effect of the IV on the DV
quasi experiment
Contain naturally occurring IV but one that already exists e.g. gender, personality
Compares different types of people with pre-existing characteristics easily
Participants cannot be randomly allocated to research conditions
Purpose of a pilot study
Small scale prototypes of a study carried out in advance of the main research to find out if there are any problems with the experimental design
Ensure time effort and money are not wasted on flawed methodology
how pilot studies can be used to improve experiments
Allow researches to Catch issues earlier on before they impact the main study
Allows researchers to check the validity and reliability of instructions To ensure they are appropriate and effective
stratified sampling
1. select the strata from three different schools
2. Calculate the ratio
3. From the ratio, find the actual numbers
4. Use random sampling to select the actual sample
s: Largely free from researcher bias, representative, because it's designed to reflect the composition of the target population and avoids misinterpretation
l: Difficult and time-consuming
normal distribution
Symmetrical bell shaped curve
(lowest) Mean median mode (highest)
skewed distribution
Asymmetric long skewers
Made as always the highest point, mean is always the lowest
what do you mean and standard deviation values suggest?
The main value suggests that people are able to concentrate better at 10am than 3pm (mean is smaller at 10am). This is because people who completed the task at 3pm took longer (385s) than people in the morning (346s). The standard deviation value suggest that was the greatest spread of scores for students in the afternoon (12.16) than the morning (7.52) (SD is larger in afternoon)
why a particular statistical test is chosen?
e.g. wilcoxon. The studies looking at a difference in a number of words recall (short vs long). A repeated measures design is used As the same participant are used in both conditions. The data (Number of words recalled) Is at least ordinal as There exists a continuous scale of measurement and there's not an equal distance between each point
why certain data types?
Nominal: Each participant will only agree in one category
at least ordinal: and falls between data are not equal / data can be ranked
interval: Intervals between data are equal
is there a significant difference at the 5% level of significant or not?
When N=10, The critical value for two tail test where P=0.05 Is 0.648 as the critical value (0.648) is greater than the calculated value (0.582), It is not significant at 5% level.
hypotheses
Directional: More, higher, past research (one-tailed)
Non-directional: There will be a difference, no past research (two-tailed)
type 1 and type 2 errors
type 1: Where the null hypothesis is rejected and the alternative hypothesis is accepted
A researcher will have concluded that the results of statistically significant when in fact they are not
False positive
If the P value was too lenient (0.1 Instead of 0.05), A type 1 error might have been made and claimed their results of statistically significant when they're not
type 2: exact opposite