e of research. Primary data is collected through direct investigation, usually in one of three ways-observation, survey and expreriment.Observation is looking at and recording what people do and how they behave can be important.Surveys usually involves asking questions of Respondents- people or organisations who reply to the questions asked. There are different ways of conducting surveys. A postal survey, where questionaires are sent through the post, or a newspaper survey, where readers are invited to fill in and return a questionnaire in a newspaper, are cheap. Telephone surveys, personal interviews and consumer's panels are more expensive because an interviewer has to be employed to interview customers. Surveys can only be useful for market research purposes if the questions asked are appropriate. A survey cannot ask every customer for their opinion. Only a fraction or Sample of customers can be surveyed. To be useful, the sample chosen must be representative of all consumers (the population). In a random sample, every potential respondent has an equal chance of being chosen, random numbers can be used to do this. To reduce the length of time, a systematic sample can be used. This is where every100th or 1000th person on a list like a telephone directory or the electoral register is chosen. In a quota sample, the sample is broken down or stratified. One problem with a quota sample is that any people that fit the description can be asked to complete the survey. A stratified random sample may get round this problem. It is a quota sample where all the respondents, the people being interviewed must be chosen at random. Experiments are experimental techniques which market researchers can use. To launch a new product is often very costly. Instead, products could be tested on groups of consumers to see whether they are acceptable and likely to sell well.-...
