Research Methods Flashcards
Research
Finding and using evidence to support a theory
- systematic way of finding things out
Research Methods
- Observations
- Self reports
- Content analysis
- Correctional studies
Aim
Identifies the purpose of the investigation
Hypothesis
A precise, testable statement of what the researchers predict will be the outcome of a study
Objectivity
Not a matter of opinion & free from bias
Reliability
If the results are consistent or not
Validity
If the results are accurate or not
Independant Variable
Variable that the researcher manipulates
Dependant Variable
Variable the researcher measures in their experiment
Operational
A full description of exactly how variables are defined, how they will be manipulated and how they will be measured
Null Hypothesis
States there’s no relationship or significant difference between the two variables being studied
Alternate/Experimental Hypothesis
States that the results aren’t down to chance, predicts the IV will have an effect on the DV
Sample
Smaller collection of units from a population used to determine truths about that population
5 TYPES
Target Population
The group of people who’s behaviour we’re interested in studying/measuring
Biased
Under/over representation of a characteristic within a sample
Generalisations
Applying results from sample back to target population
Representative
Able to accurately symbolise/reflect the target population
5 Types of Sampling
Random
Systematic
Stratified
Opportunity
Volunteer
Random Sampling
Get a list of target population and assign each a number.
- use a random generator to select number & then convert back into names
- all members have an equal chance of being selected
Systematic Sampling
Get complete list of target population & put them into a sampling frame.
- Choose nth term & select every nth person from frame
Stratified Sampling
Identify stratas and calculate the proportion of each strata in the target population
- Use random sampling to select a representative selection of participants
Opportunity Sampling
Asking whoever is ready & available at the time of study
Volunteer Sampling
PPs self-select themselves to become part of a study
Independent Groups
Half PPs are in Group A & other half are in Group B
Repeated Measures
100% PPs do Group A & then 100% do Group B
Matched Pairs
Half PPs are in Group A & other half are in Group B, but the PPs are matched accordingly to the characteristic/variable that could affect DV
Pilot Study
Trial run of a piece of research to check how it will work & allows researcher to reduce flaws
Lab Experiment
Conducted in highly controlled conditions (artificial)
Field Experiment
The IV is manipulated in a more natural setting
Extraneous Variables
Any other variables that would interfere with the IV or DV & should be controlled or removed
- 2 TYPES: Participants/Situational
Participant Variables
The differing individual characteristics that may impact the DV
Situational Variables
Any features of the experimental situation that may affect the DV
Confounding Variables
Extraneous variable type yet affects DV & IV, do change systematically with the IV
Demand Characteristics
Participants may alter their actions & responses to parts of the experiment in order ot fit the “desired” characteristics
- Can alter the results & they won’t be an accurate representation of the sample
Investigator Effects
Any unwanted influence of the investigator on the research outcome.
- It also may refer to any actions of the researcher that were related to the study’s design
Randomisation
A way of dealing with EVs
- Use of chance wherever possible to reduce the researcher’s influence on the design of the investigation
Standardisation
A way of dealing with EVs
- All PPs should be subject to the same environment, information, and experience. To ensure this, all procedures are standardised
Natural Experiment
Study of a naturally occuring situation as it unfolds in the real world
Quasi Experiment
The naturally occurring IV is a difference between people that already exists, so examines its effect on the DV
Self-Report Technique
Any method in which a person is asked to state or explain their own feelings, opinions, behaviours, and/or experiences
Questionnaire
A set of written questions used to assess a person’s thoughts and/or experiences
Interview
A ‘live’ encounter where one person asks a set of questions to assess the interviewee’s thoughts and/or experiences
Closed Questions
- Fixed number of responses
- Collect quantitative data as they limit the possible answers
- Easy to analyse but lacks detail & depth
Open Questions
- Doesn’t have a fixed range of answers
- Respondents are free to answer however they see fit
- Open questions tend to produce qualitative data
Semi-Structured Interviews
- Most likely to encounter in day-to-day
- List of pre-determined questions but interviewers can ask follow up questions or interviewee can expand
- Interviewer requires training
Mean
Add all numbers & divide by quantity of numbers
Median
Align in ascending order & pick out middle
Mode
Most common number
Range
Difference between lowest & highest number
Standard Deviation
How spread out the scores are from the average
Percentages
Figure out of a hundred
Correlation
Relationship/trend between variables (co-variables)
Dispersion
How spread out the data is
- Range/SD
Central Tendency
How close together the data is
Primary Data
First hand collection of data (collected only for purpose of research)
Secondary Data
Data collected for another purpose/research
Discrete Data
Clear, countable values between
- Discrete as units can’t be split up (decimals)
Nominal Data
Frequency/naming data, grouping people into different categories
Continuous Data
Ongoing sequence
- Measurable
- Has meaning at all points (can be decimals)
Ordinal Data
Order in measurement (ranking) that indicates some direction but doesnt show difference between interval data as the intervals aren’t uniform
Interval Data
Shows the order & exact difference/intervals between each variable & uses units of equal intervals
Line Graph
Shows change over time
- interval/continuous
Scattergraph
Shows a correlation
- Measures strength & direction of CVs
Types of Observations
- Covert
- Overt
- Controlled
- Naturalistic
- Participant
- Non-Participant
Covert Observation
The researcher is undercover & participants are unaware that they’re being watched.
- No informed consent but a debriefing takes place afterwards
HAWTHORNE EFFECT
Overt Observation
Participant is aware they’re being watched, meaning informed consent is gained
Controlled Observation
Likely carried out in a lab & is typically structured.
Reseracher is fully in control & PPs are randomly allocated to each IV group
Naturalistic Observation
Observing behaviour in the environment that it naturally occurs.
Typically involves observing spontaneous behaviours with the researcher recording what they see
Participant Observation
Where the observer is a part of the observation group (can be covert or overt).
If overt: researcher reveals the true identity & asks permission
If covert: researcher uses a fake identity
Non-Participant Observation
Researcher choses not to play a part in the observation & this may be chosen if there isn’t an opportunity to engage in a PP observation.
This may not be in person & can be done via recordings
Hawthorne Effect
If unaware, they’re less likely to change their behaviour
Types of Reliability
Test-Retest
Inter-Observer/Rater
Split-Half
Case Study
An in-depth analysis of an individual/group/event.
- Typically collects qualitative data through observations or questionnaires
- Longitudinal
Case Study Strengths
- More info & understanding of individual, so can be more confident data is correct & accurate
- Can contribute to understanding of both typical and untypical functioning
- Can generate hypotheses for future studies
- Efficient as it only takes 1 case study to disprove a theory
- Allows reserachers to study events that they couldn’t practically or ethically manipulate
Case Study Limitations
- Issues with generalisation of findings
- Subjectivity in terms of selection of material & interpretation by researcher
- Issues withh validity of personal accounts from pp & their peers - possibly due to memory decay
- Cannot establish cause & effect as they begin after the event
Three Faces of Eve
THIGPEN & CLECKLEY (1954):
- Case of Dissociative Identity Disorder where Eve White mentioned several hallucinatory experiences to a therapist where she heard voices.
- As these experiences were being recalled, Eve’s behaviour changed and became flirtatious & boisterous, referring to herself as Eve Black compared to serious and conscientious Eve White.
- Eve was studied for 100 hours over the next 14 months, finding this extra personality ws present since childhood.
- Used psychological testing & scientific backing (EEGs) to determine personalities
Ethical Issues
Arises when a conflict exists between the rights of a PP and the goals of research to produce authentic, valid and worthwhile data
BPS Code of Ethics
Quasi-legal document produced by the British Psychological Society that instructs psychologists in the UK about what behaviour is and isnt acceptable when dealing with ppts.
Build around 4 major principles
4 Major principles of BPS Code of Ethics
- Respect
- Competence
- Responsibility
- Integrity
4 Main Ethical Considerations
- Informed Consent
- Deception
- Protection from Harm
- Privacy & Confidentiality
Informed Consent
Make ppts fully aware of what the research involves (aims, rights, what their data will be used for) before they agree to be involved
Deception
Cannot purposefully midlead or withhold information from the ppts
Protection from Harm
Ppts should not be placed at any more risk of harm (psychological or physical) as they would be in everyday life
Privacy & Confidentiality
Protecting ppts’s information in the research, keeping anonimity
DATA PROTECTION ACT - right to have any personal data protected
Dealing with Informed Consent
Can gain consent through 3 different ways: Prior-general consent, retrospective consent, presumptive consent
- sign an agreement/consent form stating they’re aware of all information about the study
Prior General Consent
Ppts give consent to take part in many different studies, but unaware of which one they are taking place in
Retrospective Consent
Asking for consent after the study, or during the debrief
Presumptive Consent
Ask a group of like-minded people & generalise the consent to ppts
Dealing with Deception
Full briefing beforehand for full awareness of aims/methods/other ppts’s actions
- Give PPts the right to withdraw
Right to withdraw
PPts should be allowed to withdraw from reserach at any stage.
- if they’re being paid, they should still be paid if they leave
Dealing with Protection from Harm
- Include a 3rd party observer
- remind all ppts of their rights & everything about method
- offered therapy & counselling
- ensure ppts that their behaviour was typical and normal
Dealing with Confidentiality
Use initials/letters/numbers rather that names, distorting audio recordings & bluring any faces.
- Data should be kept in password-protected files
- All PPts should be kept anonymous
Content Analysis
Interpreting secondary material to understand individuals, where they’re studied indirectly.
- Aims to summarise and describe this communication in a systematic way so conclusions can be drawn
Content Analysis Method
- Identify what they’re looking for
- Create categories
- Read through source material and tally each time the category occurs
- Quantitative data so it allows for comparisons (like bar charts, graphs, tables)
Content Analysis Strengths
- Enables the analysis of a wide range of material
- Many communications are already in the public domain so allow for high ext validity and decreased ethical problems
Content Analysis Limitations
- Any findings are limited by the researchers expectations because categories are decided prior to the analysis
- Although researchers are trained, their interpretations of material may still be subjective
- Behaviour may be taken out of context and the researcher may attribute opinions or motivations to the source material
Peer Review Process
- Scientist conducts a study
- Scientist write about results
- Journal editor recieves article & sends it out for peer review (may need revision & review to then be sent back)
- Peer reviewers read the article & provide feedback
- If article meets the standards, it is then published
What makes psychology a science?
Follows a strict proess, the ‘golden standard’ framework of how reserach should be carried out. This is because the process provides an empirical standardised method for creating reserach, which increases the reliability & validity of findings.
Empirical/Objectivity/Hypothesis Testing/Replicability/Open to Fasification
Empirical
Study observable behaviour or cary out experiments
- Doesn’t rely on belief, carried out with careful and detailed methods so it can be replicated
Objectivity
All sources of bias are minimised and personal or subjective ideas are eliminated.
- Implies the facts speak for themselves even if different from hypotheses.
- All EVs need to be controlled to establish cause & effect
Replicability
Whether a particular method & finding can be repeated with other people or occasions to see if they’re similar.
- If a dramatic discovery is reported, but cannot be replicated by other scientists, it won’t be accpeted.
- If we get the same results over & over again under some conditions, gives confidence that results are reliable & accurate
Open to Falsification
Using scientific methods, means you’re able to prove the theory is incorrect
The Scientific Process
- Make an observation
- Ask a question
- Form a Hypothesis that answers the question, either experimental, alternative, or null
- Make a prediction based on the hypothesis
- Do an experiment to test the prediction (aims, method, and results)
- Analyse the results: Whether hypothesis is correct or not (If incorrect, go back to step 3)
- Report the results
- Peer review
Peer Review Strengths
- Increases validity, more likely to say research is accurate. Provides feedback and validity upon expert knowledge. The process is understood & accepted by majority of reserachers
Peer Review Limitations
- Anonymity is not infallible; reviewers find it hard to remain purely objective due to their education & experiences. Reviewers are highly critical of research that contradicts their own views (myside bias) & vice versa. Anonymity may be used to criticise rival researchers. Lack of research funding means researchers may struggle & favour others.
- Publication bias - editors want to publish significant ‘headline-grabbing’ findings & may prefer positive results. ‘file drawer effect’: where only statistically significant findings are published
- Burying ground-breaking research as they want to keep things as normal & may want to suppress opposition to mainstream theories
Test-Retest Reliability
Concerned with EXTERNAL reliability
- Ppts take the same test at different occasions, a high correlation between test scores indicating that the test has good external reliability.
- If the correlation coefficient is above (+0.80) it has good reliability.
- Spearman’s Rho
Split Half Reliability
Concerned with INTERNAL reliability
- Compare the performance of 2 halves of a test/questionnaire, with a close correlation between scores on both halves of the test
Inter-rater/observer/intra-rater Reliability
Inter-observer: Used to assess the degree to which different independent observers give consistent extimates of some phenomenon
Inter-rater: Same as inter-observer, but for a method other an observation (e.g. content analysis)
Intra-rater: Same reseracher is assessed for consistency of test/observation results (e.g. would they mark the same answer the same)
2 Types of Reliability
Internal: Assesses the consistency of results across items within a test
External: Extent to which a measure varies from one use to another
Ways of Improving Reliability - Questionnaires
- Ask some standardised questions
- Closed questions
- Rewrite questions so they cant be interpreted differently
Ways of Improving Reliability - Interviews
- Same interviewer
- Same training
- Structured questions
Ways of Improving Reliability - Experiments
- Standardised & controlled lab experiment
- Control EVs
Ways of Improving Reliability - Observations
- Multiple independent observers
- Operationalise categories
- Distinct categories
Ways of Improving Reliability - Content Analysis
- Operationalise categories
- Distinct categories
2 Types of Validity
Internal: Effects observed in a study are due to the changing of the IV & not other factors.
External: How much the results can be generalised to other settings (ecological, temporal, population)
Threats to the 2 types of Validity
Internal: Researcher bias, EVs, Types of ecperiment, Confounding Variables, Demand Characteristics
External: Artifical Setting, Non-representative setting/sample, era-bound, culture-bound, low mundane realism
Ways of improving the 2 types of Validity
Internal: Controlling EVs using standardisation, counterbalancing and eliminating DCs & investigator effects
External: Setting research in a more natural setting, and use random sampling to select ppts
Face Validity
Examining the content of the test to see if it looks like it measures what it is supposed to measure
Concurrent Validity
Involves comparing a new method or test with an already well-established test that claims to measure the same variable.
- A significant positive correlation should be gained between the results of the two tests
Temporal Validity
How relevant the time period is in affecting the findings.
Ecological Validity
Whether a test or method measures behaviour that is representative of naturally occurring behaviour
Population Validity
Extent to which findings from a sample can be generalised to similar and wider populations
Predictive Validity
Whether the test will predict future behaviour/performance indicated by its results
Construct Validity
Degree that a psychological test measures the abstract concept or psychological construct that it seems to measure
Ways of assessing validity
- Lie Scale
- Face Validity
- Concurrent Validity (compare performance on established & new test)
- Construct Validity (degree to which the theories have been represented)
Ways of Improving Validity - Experiments
- use a control group
- standardised procedure
- single & double blind procedures
Ways of Improving Validity - Questionnaires
- make qs easy to understand
- lie scale to assess for social desirability bias
- anonymity
Ways of Improving Validity - Observations
- covert & natural (more realistic results)
- operationalised behavioural categories
Ways of Improving Validity - Qualitative Methods
- use different methods to compensate for limitations
demonstrate interpretive validity (coherence/direct quotes)
What is science?
Pursuit and application of knowledge & understanding of the natural and social world following a systematic methodology based on evidence
Shared assumptions of the scientific approach
- Must be definable & agreed upon subject matter or method (or paradigm)
- There must be a theory construction from which hypotheses are derived and tested
- Empirical methods of investigation are used to gather info
- Science should attempt to discover gender laws & principles
- Science must be open to being tested & thus falsified
Paradigms
Collective set of assumptions about a subject & metyhod of enquiry
- Form the generally perspective of a particular subject at a point in time
What did Kuhn (1970) believe
Believed that a subject can only be a science if the majority of its workers agree with and work within a common ‘global’ theory or paradigm. He described 3 stages of scientific development
Kuhn’s 3 stages in the development of science
- pre-science (no paradigm)
- normal science (a paradigm is establiish & all researchers dedicate themselves to exploring it)
- revolution (a point is reached where so much evidence confluicts the paradigm that it is rejeced and is replaced by one which can accommodate the conflicting findings
Paradigm shift
An important change in the basic concepts & experimental practices of a scientific discipline
- can sometimes be referred to as a ‘scientific revolution’
Strengths of Psychology as a Science
- research uses scientific methods
- research is carried out through experimentation & uses many controls meaning cause and effect can be established
- ppsychology has theories that can generate hypotheses
Weakness of Psychology as a Science
- psychologists study humans which can’t be studies in the same way as chemistry & physics where people are aware theyre being studies
- even with hypotheses, lots of material is called psychology is not a science
Normal Distribution
- plotted in the form of a symmetrical bell shaped curve
- median/mode/mean are all in the middle & are very close together
- tails don’t reach zero
Skewed Distributions
Some data sets produce skewed distributions (asummetrical patterns which appear to ‘lean’ to one side)
Positive/Right or Negative/Left
Postive/Right Skewed Distributions
Most scores are low & on the left of the graph
- Leads to a long tail on the right
- mode is at the highest point, then median, then mean (left to right)
Negative/Left Skewed Distributions
Most scores are high & on the right of the graph
- Leads to a long tail on the left
- mode is at the highest point, then median, then mean (right to left)
Inferential Statistics Test
Allow us to draw conclusions about whether the results are significant or not
Significance
Research findings are sufficiently strong enough that a researcher can reject the null hypothesis and accept the alternative
- usually use p≤0.05
Steps to chose the right test
- Correlation or Difference (If correlation skip to 3)
- Experimental design - related/unrelated
- Level of measurement - nominal/ordinal/interval
- Parametric? (has to be interval)
Importance of scientific report writing
It’s a chance for psychologists to show their case findinfs of important/ground-breaking research
Scientific Report Sections
Abstract
Introduction
Method
Results
Discussion
Referencing
Scientific Report - Abstract
Short summary of overall nature of your investigation (200-300 words)
- written after the research but placed at the start
- contains :aims, hypotheses, design, variables, pps & sampling, results
Scientific Report - Intro
(600-700 words)
- Literature review (general to specific)
- Puts aims & hypotheses into context
- Key items are operationalised
- Finishing with specific aims/hypotheses
Scientific Report - Method
- Design
- Sample
- Apparatus/equiptment
- Procedure
- Ethics
Scientific Report - Results
Use of differential/inferential stats
- indicate what’s happening rather than why
- level of significance
- calculated & critical levels
- significance or not
- justify use of stats test
Scientific Report - Discussion
- explanation of results
- results are analysed in relation to studies in literary review
- highlighting any weaknesses of the study & mentioning modifications to the study if it were to be conducted again
- overall conclusion
Scientific Report - Referencing
- give recognition & credit for their work
- so others have access to your references so readers can read the referencesd articles themselves
- avoid plaigarism
Book Reference Template
Author Surname & Initial, (date of publishing), Book Title [italics], Place of Publication, Publisher
Journal Reference Template
Author Surname & Initial, (date of publishing), Title of Article, Title of Journal, Voilume number, Issue number, Start & End page numbers
Stats Test mnemonic
Carrots / Should / Come
Mashed / With / Swede
Under / Roast / Potatoes
Stats Test table
Unrelated / Related / Correlation
Nom: CHI-SQUARED / SIGN / CHI-SQUARED
Ord: MANN-WHITNEY / WILCOXON / SPEARMAN’S RHO
Int: UNRELATED T / RELATED T / PEARSON’S R
Event Sampling
Observation of targeted behaviours or specific events, recording the number of times a certain behaviour occurs.
Time Sampling
Where an observer records behaviour at prescribed interval