Sociologists using official statistics are very well aware of the quality of data it produces. These are numerical data produced by national and local government bodies and cover a wide range of behaviour including births, deaths, marriages and divorce, economical and employment records, and work and leisure. Official statistics produce a vast array of quantitative data which counts as its strength due to the objective approach quantitative data provides. In some cases, official statistics are the only source of information of sociologists' concern. Their easy accessibility and low cost positions them on preferential means of collecting information. Data collected from official statistics is often highly representative as they are based on a very large sample. However, some sociologists specifically the interpretivists do not accept the reliability and validity of official statistic data. They reject the use of statistics on the basis of their belief that statistics are the product of the meanings and taken-for-granted assumptions of those who construct them. Cicourel and Atkinson (as phenomenologists) regard officiall statistics as social creation therefore, chaleenging the objectivity they claim to provide. On other hand, conflict sociologists argue that official statistics are neither hard facts, nor subjective meanings. Instead they cosist of information which is systematically distorted by power structures in society. They believe that the collection ofthese data is manipulated by those in power so that they tend to favour the interest of the powerful. Despite all criticisms, sociologists regard official statistics useful because they are relatively an unobtrusive measure of social life which means that they avoid the interference of demand charecteristics of those being observed. Furthermore, it is generally accepted that official statistics provide useful data on the phenomena they measure.
Historical documents are another source of secondary data important to sociologists who wish to study social change which takes place over an extended period of time. Unlike official statistics, they provide rich and valuable source of qualitative data. That makes them highly valid however, a subject to criricism of losing the scientific objective approach study of social life demands according to the positivists. Other advantages of using historical sources could be that they are cheap to obtain and easily available. However, there are issues regarding the use of historical documents that limits its use. Most important ones of those disadvantages could be the authencity of data, that is, the data these sources provide might be inaccurate for sociological purposes. Another one of the weaknesses concern the problems of the credibility and representativeness of data as both of these highly depend upon the ways and priorities of the author of that document. The limitation on the side of the researcher comes when interpretating the meaning of a text, which could vary for all sociologists. Therefore, like all other sources of secondary data, historical documents too contain their own strengths and weaknesses. However, their use mainly depends upon the theoretical and practical approach of the sociologist.
Another means of attaining secondary source of data could be the use of documents. These include letters, diaries, memoirs, autobiographies, novels, newspapers, advertisements, posters, photograph and radio and tv broadcasts. Howver, our subject of concern regard the approach to carrying out content analysis, in which researchers analyse the content of documents. These may be largely quantitaive, largely qualitative, or combine both approach. Formal content analysis is one of the four approaches of analysing data. It emphasises upon objectivity and reliability and work towards the collection of systematic sample of text in order to identify different features of this text. The method is reliable because other researchers can repeat the same technique to check the findings. The data called can also be replicated to carry out comparitive studies. The simplicity and reliability of quantitative content analysis makes it appealing. However, this approach follows an assumption that the audience are simply pasive consumers of the message and no attempt is made regarding the interpretation of the message in the text. Secondly, the thematic analysis is aimed at discovering the ideological biases of those involved in the production of mass media. Critics of such studies argue that they are a non-scientific approach and tend to use examples selectively that fits the preferred interpretation of the researcher.
Conclusively, all secondary sources have their own strengths and limitations. Their use and the effectiveness of that use depends on the theoretical and practical approach of the sociologist.