Research methods Flashcards
Hypothesis
A tentative and testable explanation of the relationship between two or more variables
Types of variables
Variables is a characteristic or property that varies in amount and can be measured
IV - Variable whose effect is being studied; antecedent of DV; which the experimenter manipulates
DV - Variable that depends on IV; is a consequence
Types of Research
Correlational study - Researcher does not manipulate the IV
Quasi-experimental - Researcher does not use random assignment and lacks sufficient control over variables
True experiment - Researcher controls the levels in the IV and uses random assignment
Field study - Researcher does not interfere in what’s being studied; naturalistic observation
Types of sampling
Population - Group the researcher wishes to generalize their results to
Representative sample - Sample which is a miniature version of the population
Random selection - Every population member has an equal chance to be selected for the sample
Stratified random sample - Relevant subgroups of population are randomly sampled in proportion to size
Opportunity sample - Take whoever is available; easiest way; might not be representative of population
Volunteer sample - People who sign up themselves
Deliberate: selection of particular units that constitute sample [certain shops to interview, etc.]
Systematic: random number on the list is selected and every nth element is selected until number is secured
Cluster: Researchers divide a population into smaller groups - clusters; then randomly select among these clusters to form a sample. Used to study large popl, geographically dispersed.
Confounding variables
Unintended independent variables, a type of extraneous variable that are related to a study’s independent and dependent variables that DO effect the DV
Eg: testing if lack of exercise affects weight gain, amount of food consumed is a confounding variable
Control group
The group that does not receive the treatment
Problems in research design [and remedies]
Experimenter bias - Due to their expectation, experimenter treats groups differently [Sol: double-blinding, using standard instructions]
Demand characteristics - Cues that might suggest to the subject what researcher expects from them [Sol: single blind effect]
Placebo effect [control groups]
Hawthorne effect - Tendency of people to behave differently if they know they’re being observed [control groups]
External validity
Types of statistics
Descriptive stats - Organizing, describing & summarizing a collection of observations
Inferential - Go beyond actual observations to make inferences and provide estimates of popular characteristics
Measures of central tendency
Mode - Value of the most frequent observation in set of scores [two modes - bimodal]
Median - Middle value when observations are listed in ascending or descending order
Mean - numerical halfway point between the highest and lowest score [arithmetic average]
Measure of variability / dispersion / spread
Range - Highest minus lowest score
Standard Deviation - ‘Average’ scatter away from the mean; spread of scores around the mean
Valence - Square of SD
Percentile
Percentage of scores that fall at or below that particular score
Z-score
To calculate how many standard deviations above or below the mean your score is.
Subtract the mean of distribution from your score and divide the difference by the SD.
Negative z-score fall below the mean and positive z-score fall above the mean.
T-score
Has a mean of 50 and SD of 10, often used in test score interpretations. Easier to use because there are no negative numbers like in z-scores.
Calculate: (10 x z-score) + 50
Correlational coefficients
Correlation coefficients are used to measure how strong a relationship is between two variables.
They range from -1.00 to +1.00. The closer it is to these extreme, the more accurate the prediction. If they have a correlation of 0, then value of first variable doesn’t help predict the other.
Positive correlation - When one increases, other increases and vice versa
Negative correlation - When one increases, other decreases and vice versa
Graphical representation of correlational data is called a scatterplot.
Significance testing
Used by researchers to draw conclusions about populations based on research conducted on samples
A formal procedure for comparing observed data with a claim (also called a hypothesis), the truth of which is being assessed
Experimental hypothesis are confirmed by disconfirming the null hypothesis (by showing it is not supported by the data)
When null hypothesis is rejected, observed difference is statistically significant.
Types of significance tests
T-test - Used when you have two groups
ANOVA - More than 2 groups. Estimate how much group means differ from each other by comparing between group variance to the within-group variance using a ratio [F ratio]
Chi-square tests - Used when individual observations are names or categories. Significance tests that work with categorical data
Meta-Analysis
Statistical procedure used to make conclusions on the basis of data from different studies
Applied vs fundamental research
Fundamental - Concerned with generalization and formation of a theory
Applied - Find solutions to existing problem in society, etc.
Conceptual vs empirical research
Conceptual - related to an abstract theory; used by philosophers to develop new theories
Empirical - Based on observation; data-based
Types of quantitative research approaches
Inferential - Use data to infer relationships between populations
Experimental - Variables can be manipulated to see influence
Simulation - Creation of artificial environment where data is generated
Working hypothesis
Tentative assumption made to draw out and test its logical or empirical consequences; precise defined terms
Census inquiry
Enumeration of all items in the population for highest accuracy; impractical
Sample design
Definite plan determined before any data is actually collected for obtaining sample from given population
Ways to collect data
Observation PI Telephone interview Mailing questionnaire Schedules
Parts of research design
Sampling design - how items are selected
Observational design - conditions under which observations are made
Statistical design - how info is analysed
Operational design - techniques to carry out all procedures
Extraneous variables
Independent variables not related to the purpose of the study that MAY affect dependent variable
Confounded relationship
When DV is not free from extraneous relationships, the relationship between DV and IV is said to be confounded
Research design for exploratory studies
Used to investigate a problem which is not clearly defined; conducted to have a better understanding of the existing problem, but will not provide conclusive results; development of ideas
- Survey of literature
- Experience survey
- Insight-stimulating examples
Research design in descriptive studies
- Formulating precise objective
- Designing methods of data collection
- Selecting sample
- Collecting data
- Processing data
- Reporting findings
Principle of replication
Experiment needs to be repeated more than once to increase statistical accuracy; avoid experimental error
Principle of randomisation
Design a experiment in a way to protect it from effect of extraneous factors; better estimate of experimental error
Principle of local control
A device to reduce or control the variation due to extraneous factors and increase the precision of the experiment.
- Grouping: placing similar (homogenous) subjects into a group
- Blocking: creating different blocks for attainment of grouping
- Balancing: grouping and blocking should create designs that are balanced
Informal experimental designs
- Before-and-after without control design
- After-only with control design
- Before-and-after with control design
Formal experimental designs
- Completely randomized design (C.R. Design)
- Randomized block design (R.B. Design)
- Latin square design (L.S. Design)
- Factorial designs
Non-probability sampling
The sample is selected based on non-random criteria, and not every member of the population has a chance of being included.
Eg: Convenience sampling, snowball sampling, voluntary sampling
Probability sampling
AKA ‘random sampling’ or ‘chance sampling’; every item of the universe has an equal chance of inclusion in the sample
Systematic sampling
Sample members from a larger population are selected according to a random starting point but with a fixed, periodic interval
Eg: From a population, randomly 33 is picked as a number, so every 33rd item is taken into the sample
Between-subject design
Each subject is exposed to one level of IV.
Subjects randomly divided into two groups and 1 groups gets one level of the IV and other group gets another.
Matched subject design
Matching subjects on the basis of the variable to control; one from each pair takes part in diff conditions
- Can partly control individual differences
- No order effects
Within-subject design
Using the same subjects as participants in both groups - exposing the participants to two levels of IV; reducing the chance of individual difference
Strong coefficient correlation have ____ scatter along the y axis and ____ scatter along the fitted line.
More, less
Weak coefficient correlation have ____ scatter along the y axis and ____ scatter along the fitted line.
Less, more
2 ways to interpret test results
Norm-referenced test: Assessing individual’s performance based on how they do compared to others
Domain/criterion-referenced test: Measure how much test-taker knows about content domain; derived from test norms
Triangulation
Multiple methods of data collection and analysis to arrive at conclusive results
Qualitative Data Analysis
Narrative Analysis - understanding data from stories
Discourse Analysis – understanding that different situations create different meaning
Archival Research – using past information such as written stories, past census, personal diaries etc.
Order effects
- The order of the conditions having an effect on the participants’ behavior.
- Performance in the second condition may be better because the participants know what to do (i.e. practice effect)
- Or their performance might be worse in the second condition because they are tired (i.e., fatigue effect).
Ways to control extraneous variables
- Random allocation [standardisation, counterbalancing]
- Randomisation: conditions to be completed are randomly generated and not decided by experimenter
Standard Deviation
Larger the SD, larger the spread = more individual differences between scores
Smaller the SD, lesser the spread = more consistency in scores
P-value
How likely that the results occurred by chance; anywhere between 0 & 1
Closer to 1, more likely results came by chance
Smaller value of p, more likely to accept hypothesis & reject null
Significance
Set by probability; smaller p value more significant results
Usually set at = 0.05
Type I & II errors
Type I error: Rejecting the null hypothesis when it’s actually true [optimist error / false positive]
Type II error: Accepting the null hypothesis when it’s actually false. [pessimist error / false negative]