Information that has been personally colllected by the researcher

Data that includes official statistics, personal letters and diaries aswell as previous research by other sociologists.

Official statistics

Data that already exists in some form and that has not been personally collected by the researcher.

You can save time, money and effort

You can use secondary data for historical and comparative purposes.

Some forms of secondary data, like official statistics are highly reliable

Secondary data is always produced with the needs of sociologists in mind.

A relation who is twice removed, like a cousin

A causal connection between two or more variables where one factor is proposed to have had a definite effect on another factor.

A statistical relationship between two or more variables that expresses a level of probability. A high positive correlation suggests a strong probability of a relationship.

A statistical connection with a high level of probability which ensures with some certainty a positive relationship.

It can be useful to make comparisons of data across categories and over time

As Matveev (2002) argues, quantitative data is more reliable.

Statistical comparisons can test whether a hypothesis is true or false

Quantitative data allows us to gain an understanding of people's motivations for behaviour.

Data which includes correlations and causations

Non-numerical data that expresses the quality of a relationship

It includes, meanings, feelings and interpretations

Good quality explanations

Small groups of respondents may not be representative of anything other than themselves.

Qualitative data can provide detailed, in-depth data

Researchers can record complex reasons for behaviour and the results are more likely to show the richness of people's lives.

The researcher can ask the respondent questions about how people understand and interpret their own behaviour.

To meet the respondents and start to establish a rapport

To find flaws with the study and improve the research design.

To initiate contact and save money

To test out the research hypothesis prior to the study

They suggest that official statistics may not be valid

They consider official statistics to be social constructions

Official statistics are able to show trends and allow the research to make comparisons over time/

Official statistics may not tell the researcher anything about underlying reasons for behaviour

The wording of questions should not be ambiguous to ensure that the study is replicable

So that the questionnaire does not lead respondents into a particular answer

So that respondents are able to understand what the question is asking them

So that grammar and punctuation follow the rules of standard English.

The study can be repeated and similar information will be obtained

The interviewer can spend time probing and questioning the respondent and really make them think about their answers.

People sometimes say what they think the researcher wants to hear

Unstructured interviews can avoid researcher imposition- that is the the researcher imposing their own meaning on the interview by creating the questions.

By ensuring that all opinions are represented in the writing up of the study.

By choosing a sampling technique which will produce a sample which reflects characteristics which are typical of the wider population

By making sure that the research method can be repeated

By checking that the sample chosen reflects relevant population characteristics such as gender, ethnicity and age.