Week 1 Flashcards
6 components to the experimental research design
- current knowledge
- construct hypothesis
- design experiment
- execute experiment
- carry out statistical analysis
- interpret and report.
What is a research design
a conceptual framework
a decision matrix
a network of condition for collecting and analysing data
why is research design important
provides a smooth operation
makes research as efficient as possible
necessary precautions to reduce errors for the entire project
reliable results rely on design
what makes a research design good?
specifies sources and info needed strategic roadmap for collecting and analysing data refines timelines and costs a good design must include: - clear statement of problem - procedures and techniques (ethics) - range of processes and analysis
key criteria of a research design
- reliability
- replicability
- validity (measurement, internal, external, ecological).
define reliability
are the results from the research repeatable, and would it produce similar results. Are the measurements consistent. Are the instruments stable, e,g stopwatch.
define replicability
is the study capable of being replicated based on the criteria of reliability.
define validity
integrity of the conclusion that os drawn from the research.
measurement validity
does the measure devised for the concept reflect the concept that is supposed to measure.
internal validity
defined as the extent to which the observed results represent the truth in the population we are studying, not due to methodological errors
external validity
extent to which you can generalise the findings of the study to other situations, people, settings etc.
ecological validity.
are the scientific findings applicable to peoples everyday lives. Is an experiment conducted in a lab, replicable and applied outside the lab.
what are variables
something that can vary and unique. e.g gender, temperature
continuous variables
e.g temperature - scale
discrete variables
e.g number of symptoms of illness, only certain discrete values within the range
categorical variables
e.g marital status, ethnicity. Where values are recorded to specific category.
primary research designs (4)
- correlational
- experimental
- quasi
- cross- sectional/ survey.
describe correlational design
correlation not causation. a design where we measure the variables of interest and then see how each variable changes in relation to the changes in the other variables. - investigates relationships or association between variables.
Techniques used in correlational design
Pearsons, spearmans rho, simple or multiple regression, chi squared.
weakness of correlational design
can only test correlation not causation.
describe an experimental design (true experiment)
one of the msot widely used designs in science. Experimenter manipulates the IV to see what effect it has on the DV. Defining feature of this design is the random allocation of participants to conditions
what is the IV
value is not dependent on other variable
what is the DV
assumed to be dependent on the value of the IV
describe a quasi experimental design
involves seeing if there are any differences on the DV between conditions of the IV. No random allocation of Participants to various conditions of the IV - as not possible e.g gender. Harder to infer causation.
design techniques for experimental and quasi designs.
t-test, Mann-Whitney U, Wilcoxon, ANOVA (analysis of variance)
what is a within-participant design.
repeated measure - repeat experiment with same people 3 rtimes. same P in every condition of the IV.
between persons designs
independent measures- 2 groups e.g one men one women and see effect of alcohol. group of P/s in one condition of the IV are different from P’s in another condition
advantages of repeated measures
youu can control the inter-individual confound variables, they will have the same qualities for every condition.
you need fewer participants.
disadvantages of repeated measures
people get bored or tired, leading to order effects. and demand effects. which is where participants realise purpose of exp and conform to it.
define order effects
consequence of repeated measures, where by completing the conditions in a particular order leads to differences in the DV that are not a result of the manipulation of the IV. Can be due to practice fatigue or boredom
how do you overcome order effects?
by counterbalancing- systematically vary the order in which participants take part in the various conditions of the IV.
advantages of independent measures design
P’s less likely to get bored or tired, therefore reduces order and demand effects.
disadvantages of independent measures design
- needs more participants, so more time and money.
- lose certain degrees of control over any confounding variables
- different people bring different characteristics to the exp setting, so when randomly allocating P’s you might by chance allocate all p’s with one characteristic to the one group.