OCC2022 Class Test B - Definitions

Weeks 5-9

Convenience sampling

When participants are chosen because they are ‘convenient’ — they might be in close proximity to the researcher, etc

Non-probability sampling

Participants are chosen in a process that does not give all participants in the population an equal chance of being selected.

Probability sampling

Participants are chosen randomly from a population.

Purposive sampling

The researcher looks for cases that will be able to provide rich or in-depth information about the issue being researched.

Quota sampling

Participants are chosen according to pre-specified quotas regarding demographics, attitudes, behaviours or some other criteria.

Saturation

Little or no new data is generated from the participants, and it is believed that the sample size is adequate

Simple random sampling (SRS)

Every participant has an equal chance of selection. There are several methods of selecting a random sample from the sampling frame, such as using a statistical software package or random number tables.

Snowball sampling

Someone is identified who meets the criteria for inclusion in a study, and they recommend others who also meet the criteria

Stratified random sampling

A population is divided into groups also known as ‘strata’, and then researchers continue by either implementing SRS or systematic random sampling.

Cluster sampling

Involves selecting participants by using groups that have similar characteristics and in which participants can be found

Systematic sampling

Involves selecting every nth participant from a list of the members of the population

Dependent variable

For intervention studies, the outcome measures expected to be affected by the intervention (independent) variable.

Descriptive research

Observational studies describing and optionally comparing the characteristics or responses of a sample

Diagnostic accuracy study

A research design evaluating how well a clinical diagnostic or assessment procedure can identify the presence or absence of a health condition in a sample.

Epidemiological research

The scientific study of causes and other factors affecting the occurrence illnesses, injuries, survival or recovery among populations and groups within a population

Experimental design

An intervention study involving at least one treatment group and a control or placebo group, so the effect on outcomes with and without the intervention can be compared.

Independent variable

For intervention studies, the intervention or the treatment expected to affect the outcome (dependent) variable.

Intervention study

Research where the researchers deliberately do something with participants with the intention of causing a change. Compare with ‘observational study’.

Level of measurement

How quantitative information is recorded, according to whether only group membership (nominal scale) or ranking (ordinal scale) is identified, or whether quantities or amounts are directly measured (continuous scales). Continuous scales (interval and especially ratio) are considered more informative levels of measurement than categorical (nominal or ordinal) scales.

Meta-analysis

Statistical procedure for combining results from existing quantitative studies with measured outcomes, effectively increasing sample size and precision.

Observational study

Research where the researchers do not intervene to bring about a change but instead observe and measure events as they happen naturally. Compare with ‘intervention study’.

PICO

A way of listing the clinical population, intervention (treatment), comparison or control conditions and measured outcomes for an experiment; useful for identifying search terms for intervention studies

Reliability

In measurement, the consistency and repeatability of measurement scores over time, measurement equipment and people doing the measuring.

Systematic review

Secondary research evidence that answers a focused clinical question through a structured and rigorous synthesis of original, primary studies chosen with strict and transparent inclusion and exclusion criteria

Validity

In measurement, the accuracy of a measurement, whether a measurement system measures what it’s supposed to measure, and only what it’s supposed to measure, which depends in part on reliability.

Variable

Anything measurable, either categorically or continuously, and can differ from one observation to another, such as among different individuals, groups of people, or across time.

Deductive

Going from the general to the specific.

Inductive

Going from the specific to general

Inter-rater reliability

When two or more researchers assess the data and results and give consistent estimates of the same phenomenon.

Progressive focusing

The ability to review the research at various stages and adjust accordingly if required to improve on the study.

Reflexivity

Examining oneself as the researcher as well as the research relationship. Making the research process and researcher a focus of inquiry.

Triangulation

Improves validity of a study because it cross-validates data through verifying the research results through two or more sources, typically through the application or combination of different research methods attempting to answer the same questions.

Case

Individual members of a sample, person, place, object or event; participants in research.

Central tendency

The middle location between the highest and lowest scores on a continuous variable; relevant statistics are the mean, median and mode

Confidence interval

[for a mean, also ‘CI’] From inferential statistics, a range of values in which a population mean is estimated to occur, with a specific level of chance, conventionally 95% confidence.

Correlation

Statistical association between one measurement and another, so the results on one variable are at least partly predictable from the other. Universities select students on the basis of their school results because of a correlation, with better performance at school presumably associated with higher grades in later university results.

Descriptive statistics

Statistics that summarise or otherwise describe the characteristics of a sample, groups within a sample, or relationships between variables in a sample.

Dispersion

How spread out the scores are on a continuous variable; relevant statistics are the standard deviation, minimum, maximum and range.

Distribution

Shape of the data on a graph; whether scores are symmetrically arranged or skewed

Effect

In quantitative research, a difference or a change that happens to one measured variable, such as a measure of health, attributable to another variable such as a clinical treatment.

Generalise

Conclude or assume that results from an individual or sample will apply to a larger group or population

Hypothesis testing

Using inferential statistics to estimate the probability that an effect observed in the sample is consistent with an effect, or no effect, occurring in the population

Inferential statistics

Statistics that attempt to generalise results from a sample to the wider population from which the sample was taken.

Minimum importance difference

The least amount of change in outcome measurements for a treatment or other intervention to be considered beneficial and worthwhile in clinical practice.

P value

In hypothesis testing with inferential statistics, the probability of the effect observed in the data if there were no effect in the population. Conventionally, a p value of less than .05 (5%) works as evidence of a non-zero effect in the relevant population.

Population

In statistics, all possible cases that could be sampled. In research, the clinical population comprises the entire set of people with the same clinical characteristics as in a random sample from that population.

Score

One case’s result on one variable (e.g. one person’s height).

Statistically significant

An observed effect in a sample is considered large enough to be unlikely (p value less than .05) as a chance result, so it’s interpreted as representing a genuine effect in the population rather than a chance effect in the sample.

Convergent study

Qualitative and quantitative arms of the study are run concurrently but separately and then the results are ‘converged’ to enable the researcher to compare and contrast quantitative results with qualitative findings, or expand quantitative results with qualitative data.

Embedded study

A combination of quantitative and qualitative methods is used throughout a study

Mixed methods research

The planned mixing of quantitative and qualitative components within a single study.

Multiphase study

Different study designs are used to answer the same question in a population that has inherently different sizes.

Qualitising

Converting quantitative data into qualitative data.

Quantitising

Transforming qualitative data into quantitative data

Sequential study

A qualitative study is conducted first, followed by a quantitative study (or vice versa)

Target population

The population to which the researcher ideally wants to generalise study results to

Accessible/Study population

The population to which the researcher has access to

Sample

Subset of the population that is ideally representative of the population

Participants/Respondent/Subject

A specific individual participating in a study

Sampling technique

The specific method used to select a sample from a population

Representation

The extent to which a sample or subgroup is representative of the population

Generalisation

The extend to which the results of the study can be reasonably extended from the sample to the population

Sampling error

The chance occurrence that a randomly selected sample is not representative of the population due to error inherent in the sampling technique

Measurement process

The process of quantifying information or measurement of a primary concern

• The translation of observations into numerical values or numbers

Conceptualisation

Identifying and defining concepts that are to be measured

Operationalisation

• Example

Determining the operational definition of what those concepts are

E.g. when looking at a child's visual and motor integration skills you may look at their hand writing

Level and scaling

Determining the level of measurement and selecting the appropriate scaling technique

Test-retest reliability

The consistency between time one and time two in the same group of participants who will complete the assessment tool

• Assumed that they will get the same score

Internal consistency

Measures how well different items on a survey or test that measure the same general construct produce similar scores

Equivalence reliability

Measures the degree to which two or more different versions of an instrument, or multiple indicators of the same construct, yield consistent results

Split form reliability

A method used to measure the internal consistency of a test or survey by dividing a single test into two halves and comparing the scores

Alternative reliability

2 forms of the same scale are completed by the same group of participants and then are correlated with each other

Parallel form reliability

Measures the correlation between 2 equivalent versions of a test

• 2 different assessment tools or sets of questions designed to measure the same variable, factor or trait

Rater reliability

The comparison of 2 sets of raters evaluating the same test performances

Intra-rater reliability

Same group of participants are evaluated by the same assessor on 2 different occasions

Content validity

Refers to the degree to which the indicator reflects the basic content of the phenomenon or domain of interest

Criterion-related validity

Demonstrates a correlation or relationship between the scale under consideration and another instrument that has been shown to be accurate

Concurrent validity

Measures the same or similar factor or construct to determine whether assessments, measures and scales are measuring the same thing

Predictive validity

Used when the purpose of an instrument is to predict the future performance of an individual

Discriminant validity

Seeks to show 2 measures that are not theoretically supposed to be related, are not related

Structural validity

The degree to which scores of a questionnaire are an adequate reflection of the dimensionality of the construct, attribute or factor being measured

Dimensionality

Refers to the structure of a specific phenomenon

Uni-dimensionality

 Refers to one dominant phenomenon or latent variable

External validity

The extent to which generalisation of findings from a study to other situations, people, settings and measures can be done

Population validity

Refers to whether a researcher can reasonably generalise their findings from their study sample to a larger group of people

Ecological validity

Refers to whether a researcher can reasonably generalise findings of a study to other situations and settings in the real world

Internal validity

The extent to which the observed results from a study represent the truth in the population a research is studying

Known-groups validity

Demonstrated when a test or questionnaire can discriminate between 2 groups of respondents known to differ on the variables of interest

Credibility

Relates to the evaluation of the truth value within the research

Transferability

How the researcher demonstrate that the study's findings are applicable to other contexts

Confirmability

Ensures that the researcher's interpretations and findings were clearly derived from data, requiring the research to demonstrate how conclusions were reached

Dependability

Extent that the study could be repeated by other researchers and that the findings would be consistent

Audit trail

Provides readers with the decisions and choices made by the researcher regarding theoretical and methodological issues throughout the study, which requires a clear rationale for such decisions

Memos

The theorising write up of ideas about substantive codes and their theoretically coded relationships as they emerge during coding, collecting and analysing data, and during memoing

Reflexive journal

Used to record and document the daily logistics of the research, methodological decisions and rationales