Pupils

a) When making a research among pupils it is important to consider some problems which sociologists have to avoid such as sensitivity of children: they might be shy and not answer questions for example in group interviews, that are interviews based on a group of people and where the interviewer asks questions and respondents answer in turn. So sociologists are more likely to choose unstructured interview where the interviewer can develop rapport that is friendly and trusting relationship and thereby collect more valid and in depth data. However,
Willis, for example, used in his study of ‘lads’ group interviews which allowed pupils to talk freely about their views and values and this gave him opportunity to understand reasons of creating counter school culture. However the researcher could find it difficult to interpret interviewees answers and this might lead to the interviewer bias.
Questionnaire is another convenient type of research which might help to collect data among pupils about some sensitive topics which they might not discuss orally, however this type of research does not guarantee validity of data since pupils might not respond correctly and they just may not understand a question. Also questions might mean for every person something special, so they can answer for the same question in different ways. This makes quantifying and making statistics more difficult. Marxists Bowles and Gintis in their research used self-completion questionnaires for employers and students to compare which personal traits are valued by students themselves and which personal traits are valued by employers in their workers. They could compare and find a correlation between these personal traits to explain and support their Marxist approach. Questionnaires gave opportunity for the researchers to use a large sample and use comparative method. However, questionnaires may be biased since some students could try to show themselves in favourable light and also they could misunderstand some questions.
When researcher wants for example find out how many pupils like mathematics, it is difficult to decide which sample he has to choose. Because if it is a random or systematic sample, it doesn’t mean that it will be representative, because a lot of pupils might be from science ability group and the result of the research will not be valid. Stratified samples are more useful since they give opportunities for researchers to divide the sample into strata such as age, type of school, type of ability group, etc. and make the sample more representative.