Embedded Files
Scientific investigation methodologies
Scientific investigation methodologies
Case study
A deep dive into a specific person / activity / behaviour
Correlational study
Observation of events/behaviours that haven't been controlled, to understand relationship and possible association between variables.
Simulation
Using a model to study behaviour of a real/theoretical system.
Classification / identification
Arrangement into sets, and sorting into those sets respectively
Fieldwork
Observing/interacting with environment, usually to determine correlation. Conducting with a range of other methods.
Modelling
Construction and manipulation of a physical or conceptual model, that represents concepts in order to better understand.
Controlled experiment
An experimental investigation of relationship between 1< IVs and DVs , controlling all other variables. Could use control groups too.
Literature review
Collection and analysis of secondary data in order to answer a question, or provide background information for an investigation.
Product, process or system development
Design/evaluation of a process or system to meet a human need. May involve technology.
Data and measurement
Data and measurement
Accuracy
How close it is to the true value. Not quantifiable, so just 'less' or 'more'.
Precision
How close a set of measurements are with each other. Unrelated to true value.
Repeatability
Closeness of results to those of successive measurements. In nearly exact same circumstances.
Reproducability
Closeness of results to those of successive measurements. In different circumstances. (observer, measuring instrument, location, etc.)
True value
The value/s if it could be measured perfectly.
Validity
How well the results represent findings outside of the study.
Internal = Measures what it's supposed to, appropriate design etc.
External = Results can be applied to individuals in different setting
Ethical concepts
Ethical concepts
Beneficience
Maximising benefits, minimising risks.Integrity
Honest reporting of results.Justice
Fairness.Non-malificence
Avoiding causing harm, at least not disproportionate to the benefits.Respect
Valuing wellfare, liberty and autonomy, etc. of both individuals and collective.
Ethical guidelines
Ethical guidelines
Confidentiality
Privacy and protection of participant's personal information.Debriefing
After experimentation, participants understands and has questions answered. Essential for deception.Informed consent
Participants are informed and consent to the experiment, including potential risks. Voluntary written consent.
Use of deception
Only permissable if informed consent disrupts results. Must debrief.Voluntary participation
No coercion or pressure, just free choice.Withdrawl rights
Participant can discontinue involvement at any time, can remove results.
Errors
Errors
Random
Affects precision of measurements.
Unpredictable variations.
Effects can be reduced by making more measurements, increasing sample size, or refining measurement method/technique.
Systematic
Affects accuracy of measurements.
Affects every true value by the same amount.
Often caused by instrumentation error or environmental interference.
Personal
Mistakes, miscalculations, observer errors.
Other
Other
Uncertainty
All measurements are subject to uncertainty, especially in psych where psychological concepts are used. Should be minimised as much as possible.
Outliers
Distant point from other data. May occur by chance. Can affect results, especially mean.
Variables
Variables
Controlled
Researcher controls/holds constant in an experiment, so that they do not affect the DV.
Independent (IV)
Quantities are controlled, selected or changed by researcher, which is assumed to have an affect on DV. Plotted as the horizontal axis on a graph.
Dependent (DV)
Is measured by researcher, to see the effects of changes in the IV.
Plotted as the vertical axis on a graph.
Extraneous
Any variable that isn't the IV, but may affect the DV. Should be controlled, or monitered, so they don't affect the internal validity and become confounding.
Confounding
Variables that have aleady affected the results (DV), but aren't the IV. May be an extraneous variable that wasn't controlled. They interfere with internal validity.
Experimental designs
Experimental designs
Between-subjects design
Different participant groups receive different levels of the IV.
Within-subjects design
(A.k.a. repeated measures) One same group of participants receives all levels of the IV.
Counterbalancing
Is where different participant groups complete the same conditions but in different orders, to eliminate any order effects.
Google Sites
Report abuse