Statistics is the study of the collection, organization, analysis, interpretation and presentation of data. There are variety of different types of statistics, hard statistics are numerical data that can not be manipulated, soft statistics are numerical data that can be manipulated, official statistics are produced by formal institutions whilst unofficial are carried out by informal institutions. By definition, statistical data is quantative data which increases its reliability. Using Statistics in research has a variety of different advantages.
This kind of research is cheap and easily accessible. Also, as not much time is spent on primary research, analysis will be fairly quick so the results are fairly up to date. The ready availability of official statistics mean the researcher does not have to spend time and money collecting his / her own information. Unlike in Qualitative research ...view middle of the document...
You cannot go back in time to collect data so you have to use what has already been collected.
Despite the fact that Statistical data makes a piece of research highly reliable it also has it’s weaknesses. The definitions used by the collector of statistics may not be the same as those used by the sociologist. For example, we need to ensure that the official definition of such concepts as "crime", "unemployment" and "class" is the same as our own. Also the basis for the collection of statistics may change over time. This is significant if your research is concerned with, for example, comparisons of employment / unemployment levels between the present and the past. Between 1980 and 1990, for example, the British government changed the way it defined unemployment so the official statistics were compromised.
Quantitative data gives no indication of people’s meanings or experiences. When using statistics, therefore, we need to be aware of two points: Firstly, any statistical account will represent a "snapshot" of social interaction as it was at the moment the statistics were compiled. Secondly, and perhaps more importantly, the statistical data we use may only represent a partial picture of people’s experience. This means that statistical research lacks validity.
Statistics, by definition, do not tell us anything about why someone does something (such as
committing a crime, getting married or divorced, etc.). In other words, we do not get any real
depth of information about people’s behaviour from statistical sources. A further problem, in terms of validity, is that statistics don’t always measure what they claim to measure: e.g. “Crime statistics” do not actually measure the real level of crime in this country - they actually only measure “crimes known to the police”. “Unemployment statistics” only count those people “who are available for work” on a particular day.
To conclude although statistical analysis adds high reliability to research and can be easy to carry out but consequently the research will lack validity and depth.