PSYC2009 Study Guide - Final Guide: Observational Error
Uncertainty and Psychological Research
Why we need statistical methods?
• To describe what we have measured.
• To understand and deal with risk.
• To enable us to distinguish real patterns from ones due to randomicity.
• To enable us to make inferences to populations.
Science and Paths to Knowledge
• Scientific communities have norms that make them less likely than other communities to fall
prey to a misleading authoritative consensus:
o Universalism – you do’t hae to e born right to be a scientist.
o Organised scepticism – everybody is supposed to hold all theories as provisional.
o Communalism – sharing ideas and information.
o Disinterestedness (or impartiality) – scientists are not prejudiced towards particular
theories.
o Honesty
• Scientific methods are grounded in empirical methods, based on first-hand experiences of the
world. Assumptions:
o Empiricism – a doctrine that ascribes superior truth-status to things that have been
directly observed or manipulated over things that cannot be observed or manipulated.
Researh ad theory are to e judged o their o erits regardless of the sietist’s
gender, ethnicity, etc. eg: blind peer review.
o Rationality – involves adherence to a system of reasoning (usually standard logic).
o Rationalism – a faith that rationality is the best guide to decision making.
o Statistics and probability are a combination of empiricist and rationalist ideas.
Uncertainties in Research
• All research has three common characteristics:
o Dealing with ignorance and uncertainty.
o Disputation and conflict within a framework shared by other researchers.
o Learning and discovery through errors.
• The kinds of uncertainty dealt with in research include:
o Distortion: systematic error
- Confusion (qualitative) - mistaking one thing for another.
- Inaccuracy (quantitative) - systematic miscalibration.
o Incompleteness: missing or indeterminate information
- Absence of information.
- Uncertainty - indeterminate information.
➢ Probability and statistical uncertainty
➢ Ambiguity - multiple possible meanings or a range of values.
Sources of uncertainty arise at each of its five stages:
1. Topical uncertainty - what is it that you are trying to investigate.
2. Investigative uncertainty – concerns research questions and hypotheses.
3. Methodological uncertainty – the design of the study and whether it will answer our
questions or test our hypotheses.
4. Statistical uncertainty - what is the best way to analyse and summarise this data.
5. Interpretive uncertainty – when we are unable to decide between competing
interpretations for what we have found.
Types of Research and Uncertainty
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Why we need statistical methods: to describe what we have measured, to understand and deal with risk, to enable us to distinguish real patterns from ones due to randomicity, to enable us to make inferences to populations. Assumptions: empiricism a doctrine that ascribes superior truth-status to things that have been directly observed or manipulated over things that cannot be observed or manipulated. Confusion (qualitative) - mistaking one thing for another. Ambiguity - multiple possible meanings or a range of values. Sources of uncertainty arise at each of its five stages: topical uncertainty - what is it that you are trying to investigate, methodological uncertainty the design of the study and whether it will answer our. Investigative uncertainty concerns research questions and hypotheses. questions or test our hypotheses: statistical uncertainty - what is the best way to analyse and summarise this data. Interpretive uncertainty when we are unable to decide between competing interpretations for what we have found.