Concise Glossary of Research Methods

A

• Action Research

Action research is “learning by doing” – a group of people identify a problem, do something to resolve it, see how successful their efforts were, and if not satisfied, try again. Action research is known by many other names, including participatory research, collaborative inquiry, emancipatory research, action learning, and contextural action re­search, but all are variations on a theme.

• Attrition

A reduction in the number of participants during the course of a study. If more participants withdraw from one group than another group, this can introduce bias and threaten the internal validity of the research.

• Attribution

The association or disassociation of a particular attribute with a par­ticular population unit.

B

• Bias

A loss of balance and accuracy in the use of research methods. It can creep into research via sampling, while interviewing, in the design of questions, or in the way data are analysed and presented. Bias means that the research findings will not be representative of, or generalisable to, a wider population.

• Biographical Research

Primarily qualitative, and includes gathering/ using data in the form of diaries, stories and life histories.

C

• Categorical variable

A variable with discrete values (e.g. a person’s gender or a person’s marital status).

• Causal relationship

A relationship where variation in one variable causes variation in another.

• Chi-square

Chi-square is a family of distributions commonly used for significance testing. The most common variants are the Pearson chi-square test and the likelihood ratio chi-square test.

• Coded data (coding)

Refers to a way of recording material at data collection, either manually or on computer, for analysis. The data are put into groups or categories, such as age groups, and each category is given a code number.

• Cohort study

A cohort study is one in which subjects who presently benefit from an activity are followed over time and compared with another group who are not benefiting from the activity or intervention under investiga­tion.

• Confidence interval

A confidence interval identifies a range of values that includes the true population value of a particular characteristic at a specified probability level (usually 95%). (See Statistical Analysis).

• Confidence level

The confidence level tells you how sure you can be that this inference is correct (See section on Statistical Analysis).

• Construct

Something that exists theoretically but is not directly observable.

• Continuous variable

A variable that can take on an infinite range of values along a specific continuum (e.g. weight, height).

• Controlled variables

Researchers may control some variables in order to allow the research to focus on specific variables without being distorted by the impact of the excluded variables.

• Correlation coefficient

A measure of the degree of relationship between two variables. A corre­lation coefficient lies between +1 (indicating a perfect positive relation­ship), through to 0 (indicating no relationship between two variables) to -1.0 (a perfect negative relationship). (See Statistical Analysis sec­tion for more details).

• Cross-tabulating

The process of analysing data according to one or more key variables.

A common example is to analyse data by the gender of the research subject or respondent, so that you can compare findings for men with

findings for women. Also known as cross-referencing. (See Statistical Analysis section for more details).

• Cross-sectional research

Cross-sectional research is used to gather information on a population at a single point in time.

D

• Data saturation

The point at which data collection can cease, when data becomes repeti­tive and contains no new ideas, the researcher can be reasonably confi­dent that the inclusion of additional participants is unlikely to generate any new ideas. (Sometimes simply referred to as saturation.)

• Demographics

Information about a population sample that includes data such as age, sex, social class, number of children, etc.

• Dependant variables

In a research project which seeks to establish cause and effect between variables, the potential causal variable is known as the independent variable, and the variable(s) where effects are under scrutiny is depend­ent.

• Descriptive statistics

Statistical methods used to describe or summarise data collected from a specific sample (e.g. mean, median, mode, range, standard deviation). (See Statistical Analysis section for more details).

• Determinism

The belief that everything is caused by specified factors in a predictable way rather than haphazardly; a key assumption within the positivist paradigm.

• Deviation

The difference of a score from the mean.

• Discrete variable

A variable which can only have whole numbers (integers).

E

• Emancipatory research

Conducted on and with people from marginalised groups/communities and is conducted largely for the purpose of empowering members of that community and improving services for them.

• Empirical research

Research conducted ‘in the field’, where data are gathered first hand. Case studies and surveys are examples of empirical research.

• Ethnography

Uses fieldwork to provide a descriptive study of human societies.

• Evaluation

A form of research used to assess the value or effectiveness of social care interventions or programmes.

• Experimental group

The group that receives the treatment is called the experimental group and the other group is called the control group.

• Extraneous variables

These are variables that influence the outcome of research, though they are not the variables that are actually of interest. These variables are undesirable because they add error to an analysis.

F

• Facilitator

A facilitator is someone who skillfully helps a group of people under­stand their common objectives and assists them to plan to achieve them without taking a particular position in the discussion.

• Factor

Anything that contributes causally to a result; “a number of factors determined the outcome”.

• Feminist research

Research into the relationship and understanding of the social con­structions of gender.

• Filter

When only a section of the total sample are required to answer the question.

• Frequency distribution

A visual display of numerical values ranging from the lowest to the highest, showing the number of times (frequency) each value occurs.

• Frequency tables

A set of data, which provides a count of the number of occasions on which a particular answer/response has been given across all of those respondents who answered the question.

G

• Gaussian distribution

A theoretical frequency distribution for a set of variable data, usually represented by a bell-shaped curve symmetrical about the mean. Stat­isticians and mathematicians uniformly use the term “normal distribu­tion” while physicists sometimes call it a Gaussian distribution.

• Generalisable

In technical use, has a meaning of how results from a sample can be generalised to a greater or lesser extent according to the outcome of statistical tests of significance.

H

• Hard data

Precise data, like dates of birth or income levels, which can reasonably be subjected to precise forms of analysis, such as statistical testing.

• Hypothesis

A theory or prediction made about the relationship between two vari­ables.

I

• Independent variables

The causal variable is known as the independent variable, and the variable(s) where effects are under scrutiny are dependent variables.

• Inference

The reasoning involved in drawing a conclusion or making a logical judgment.

• Inferential statistics

Statistics that allow a researcher to make inferences about whether relationships observed in a sample are likely to occur in the wider popu­lation from which that sample was drawn.

• Informed consent

The process of obtaining voluntary participation of individuals in re­search based on a full understanding of the possible benefits and risks.

• Interval level

See confidence level

• Interval variable

An interval variable is similar to an ordinal variable, except that the intervals between the values of the interval variable are equally spaced.

N

• Likert scale

A method used to measure attitudes, which involves respondents indi­cating their degree of agreement or disagreement with a series of state­ments. Scores are summed to give a composite measure of attitudes.

• Literature review

Brings together a range of information on a topic to develop an aware­ness of the current state of knowledge in the subject. It is commonly used to set the scene for introducing new research or a new perspective on the research.

• Longitudinal research

A research process, which is repeated on several occasions over a period of time, as far as possible replicating the chosen methodology each time. The key aim of such research is to monitor changes over time.

M

• Macro

A macro is a rule or pattern that specifies how a certain input sequence (often a sequence of characters) should be mapped to an output sequence (also often a sequence of characters) according to a defined procedure. Used in computer programs to conduct repetitive tasks.

• Mean

The average of your sample, computed by taking the sum of the indi­vidual scores and dividing them by the total number of individuals. (2,6,9,32,74 = 123/5 = 24.6).

• Median

If you rank the observations according to size, the median is the obser­vation that divides the list into equal halves. (2,6,9,32,74 = 9).

• Method/Methodology

While ‘method’ describes what you as a researcher have done, method­ology is about your reasons for doing it.

• Meta-analysis

A statistical technique for combining and integrating the data derived from a number of experimental studies undertaken on a specific topic.

• Mode

The observation that occurs most frequently.

• Multivariate analysis

Techniques used to analyse data that arises from more than one vari­able.

• Naturalistic paradigm

This paradigm assumes that there are multiple interpretations of reality and that the goal of researchers working within this perspective is to understand how individuals construct their own reality within their social context.

• Nominal scale

A nominal scale is one that allows the researcher to assign subjects to certain categories or groups. For example, with variable of gender, respondents can be grouped into two categories male and female. These two groups can be assigned code numbers 0 and 1.

• Normal distribution

A theoretical frequency distribution for a set of variable data, usually represented by a bell-shaped curve symmetrical about the mean. Stat­isticians and mathematicians uniformly use the term “normal distribu­tion” while physicists sometimes call it a Gaussian distribution.

• Null Hypothesis

The prediction that there is no relationship between your treatment and your outcome.

O

• Ordinal Variable

Variables with an ordered series, e.g. “very poor, poor, no opinion, good, very good”. Numbers assigned to such variables indicate rank order only, the “distance” between the numbers has no meaning.

P

• Panel studies

Panel studies measure the same sample of respondents at different points in time.

• Paradigm

A philosophical and theoretical framework of a scientific school or discipline within which theories, laws, and generalisations and the experiments performed in support of them are formulated.

• Parameter

A quantity (such as the mean or variance) that characterises a statisti­cal population and that can be estimated by calculations from sample data.

• Phenomenology

A research methodology which has its roots in philosophy and which focuses on the lived experience of individuals.

• Population

See research population

• Positivism

A paradigm that assumes human behaviour is determined by exter­nal stimuli and that it is possible to use the principles and methods traditionally employed by the natural scientist to observe and measure social phenomena.

• Predictive research

Concerned with identifying indicators of future behaviour or demand in a population on the basis of the current behaviour and demands of a sample. Predictive techniques use a number of statistical approaches.

• Primary source

A primary source is a document, speech, or other sort of evidence writ­ten, created or otherwise produced during the time under study.

Q

• Qualitative

Concerned with a quality of information, qualitative methods attempt to gain an understanding of the underlying reasons and motivations for actions and establish how people interpret their experiences and the world around them. Qualitative methods provide insights into the setting of a problem, generating ideas and/or hypotheses.

• Quantitative

As the name suggests, is concerned with trying to quantify things; it asks questions such as ‘how long’ or ‘how many’. Quantitative methods look to quantify data and generalise results from a sample of the popu­lation of interest. They may look to measure the incidence of various views and opinions in a chosen sample, for example.

R

• Random sample

A sample of a population where each member of the population has an equal chance of being in the sample.

• Range

A measure of variability indicating the difference between the highest and lowest values in a distribution of scores.

• Ratio scale

Ratio scales are like interval scales except they have a zero point. A good example is height or temperature. These have a scale with an absolute zero. Thus, a height of 2 metres is twice as high as a height of 1 metre.

• References

A reference is a formal system for drawing attention to a literature source, usually published, both in the report itself and in the bibliogra­phy or reading list at the end of the report. There are two main meth­ods of referencing articles in journal and book publications. These are known as the Harvard (author-date) and Vancouver (author-number) reference systems.

• Relevance

Is about the closeness with which the data being gathered feeds into the aims of the study.

• Reliability

The extent to which the same result will be repeated/achieved by using the same measure.

• Research plan

This is the researcher’s guidebook for the project, and the yardstick against which the various stages of progress can be judged. It states the outputs to be delivered and the timescale.

• Research population

The total number of potential subjects for your research.

• Respondent

An individual or organisation that responds to research questions.

• Response rate

The proportion of people asked to take part in research who actually take part.

S

• Sampling

The process by which you reduce the total research population for a research project to a number which is practically feasible and theoreti­cally acceptable (the sample).

• Sampling: non random

Non random sampling means that the principle of randomness has not been maintained in the selection of a sample. Often it involves structured sampling whereby the sample group is carefully matched to the overall population on key variables.

• Sampling: random

Each individual is chosen entirely by chance and each member of the population has a known, but possibly non-equal, chance of being included in the sample.

• Sampling: simple random sampling

Each individual is chosen entirely by chance and each member of the population has an equal chance of being included in the sample.

• Sampling: stratified sampling

A stratified sample is obtained by taking samples from each stratum or sub-group of a population.

• Sampling frame

The listing of the accessible population from which you’ll draw your sample is called the sampling frame.

• Secondary source

A secondary data source is that collected by other people, so for exam­ple the Census.

• Significance level

A significance level indicates the probability that an observed differ­ence or relationship would be found by chance.

• Soft data

A characteristic of qualitative research. Data such as people’s ideas and opinions.

• Stakeholders

People with an interest in the research being undertaken. For example, if the research is about an imitative which has occurred at a particular geographic level, then the corresponding level of governance would be a stakeholder e.g. a local authority or a regional development agency.

• Standard deviation

A descriptive statistic used to measure the degree of variability within a set of scores.

• Statistical analysis

Statistical analysis refers to a collection of methods used to process large amounts of data and report overall trends.

• Statistical significance

Tests of statistical significance, of which the best known is probably the Chi-square, which is a measure of probability. Where a research sample has been used, it is important to know, whether the findings are valid or came about by chance.

• Statistical tests

See section on statistical analysis for a description of the most common statistical tests.

• Survey design

Survey design covers the definition of all aspects of a survey from the establishment of a need for data to the production of final outputs.

T

• Tabulations

A set of data, which provides a count of the number of occasions on which a particular answer/response has been given across all of those respondents who tackled the question.

• Textual analysis

Used in analysis of secondary source data and also in qualitative research. It involves working on a text in depth, looking for keywords and concepts and making links between them. The term also extends to literature reviewing. Increasingly, much textual analysis is done using computer programs such as NVivo, ATLAS.ti, NU*DIST.

• Trend studies

Trend studies establish a pattern over time to detect shifts and changes and are valuable in describing long-term changes in a population.

• Triangulation

A multi-method approach, using different methods in order to focus on the research topic from different viewpoints and to produce a multi­faceted set of data. Also used to check the validity of findings from any one method.

• Type I Error

Rejecting the null hypothesis when it is true.

• Type II Error

Accepting the null hypothesis when it is false.

U

• Universe

The term universe is used to denote whatever body of people is being studied.

• Validity

Concerns the extent to which your research findings can be said to be accurate and reliable, and the extent to which the conclusions are war­ranted.

• Variable

Any factor, which may be relevant to a research study. For example the age and gender of respondents would be variables. See also Standard Variables, Dependent/Independent variables, and Controlling vari­ables.

• Variation (variance)

A measure of the spread of the variable, usually used to describe the deviation from a central value (e.g, the mean).

• Weighting

the process of weighting involves emphasising some aspects of a phe­nomenon, or of a set of data – giving them ‘more weight’ in the final effect or result.