Research methods key terms

primary data

information collected by the people who are using it. The researcher’s own observational records or questionnaires are further examples.

E.g. Jan Pahl used interviews to identify how families made decisions

secondary data

information collected by other people before it is then used by a sociologist. Media content or someone’s personal letters are other examples.

For example, the Census is a national survey carried out by the government every 10 years, the statistics of which are available to access online.

qualitative data

information that is not numerical and is often in the form of written descriptions or images. This data aims to reflect human experience and the meaning behind this. It can often be descriptive and detailed.

quantitative data

information in the form of numbers and statistics, linked to counting and measuring aspects of behaviour. This data can be used to identify trends and patterns, and is often needed to test a hypothesis suggesting a causal or non-causal relationship between two co-variables.

valid

when data is a genuine, accurate reflection of what is being studied. demographic data is often seen as being more ? as few go unreported and unrecorded.

official crime statistics produced by the police are often seen as being less ? as many crimes go unreported and unrecorded for a number of reasons

reliable

when a study can be repeated and similar results are recorded.

collecting data by a questionnaire is seen as being more ? as the method can be standardised and the researcher has less influence on the process compared with interviews which are seen as less ?

demographic data

official numbers of births or deaths in the population

ethics

issues concerning the morality of research, or the right and wrong way to undertake research.

There are guidelines to help professional sociologists. Protecting both the researchers and the participants during the research is a main example of this

objectivity

when research is conducted in an unbiased and impartial way. Additionally, the researcher does not allow their personal opinions or values to influence the research

generalisability

when data can be applied to the whole population.

Studies of educational attainment based solely on boys can’t be ? to the whole education system because they do not include girls.

representativeness

when a sample reflects the characteristics of the target population.

Most research has to be carried out using a smaller group/sample taken from the population you are interested in.

operationalisation

when in research a concept is clearly defined so that they can become measurable.

Many sociological concepts are hard to measure (gender, inequality, a stable family life, educational success). “Does Britain have a stable family life?” - what does stable mean? Amount of arguments, financially stable, relationships within the family ect.

triangulation

when a sociologist uses more than one method in their study.

Often to compare both quantitative and qualitative data, or to compare the data generalised from different methods or sources. This can maximise both validity and reliability.

methological pluralism

when a sociologist uses more than 1 method in a study but collects both qualitative and quantitative data