Information on any field, when expressed qualitatively and/or quantitatively, is called data and they are usually classified into two main categories—primary and secondary data, depending on their origin or source.
Information when collected directly from the desired field for the use of a predetermined purpose is called the primary data e.g., heights and weights of the students in a class.
It can thus be used with much confidence because of its direct nature of collection. The primary data are usually published by the authorities who are directly responsible for its collection. However, it requires enough manpower, time and money to make the process successful.
Secondary data, on the other hand, are the data already collected by an individual or an agency for a specific purpose and used afterwards for a number of other purposes. The data so obtained from that origin are then conveniently compiled up by an investigator to meet his or her own specific needs. Such data are usually collected by an authorised agency for one particular purpose and then it is officially published and again used by other individuals or agencies to serve their own purposes and in that sense these data are not original in nature.
For example, we often’ use Census (data collected every after ten years with the government effort) reports to accomplish, support and supplement our studies. These are, no doubt, secondary data used by the researchers of various fields.
The most important and remarkable advantage here is that it requires less manpower and time and as a result less costly to complete the entire procedure. But, in practice, it frequently contains a number of errors due to erroneous transcription, faulty rounding-off (up or down), etc., and, therefore, less dependable in nature for the researchers.
The investigators and the scholars working with them should therefore be much careful while using them in their own fields. It is, therefore, quite clear that the data which are initially primary in nature at the origin for one use becomes secondary in qualities and character for other uses.
Therefore, the distinction between primary and secondary data is one of degree only. A particular data may be primary in the hands of data collecting authority but may be secondary to other people using them afterwards. Prof. H. Secrist in this context says: “The distinction between primary and secondary data is largely one of degree. Data which are secondary in the hands of one party may be primary in the hands of another.”
However, the method of collection of primary data and secondary data must not be identical in nature because, in the former case data are collected originally while, in the latter case, data are to be taken up in the nature of compilation.
There exists various methods for the collection of the primary and secondary data. Choice of the exact method depends largely on the nature, object and scope of statistical investigation.
Advantages of Primary Data:
It is usually preferable to use primary data because of the following reasons: First, it generally contains a detailed description and information of the definition for the terms used. Secondly, since secondary data are second-hand data or ‘finished products’, an element of error may creep in afterwards.
This may then give some misleading information. Primary data cannot have such errors. Thirdly, in the primary data precise definition of the terms used is given and the scope of the data is clearly mentioned. Finally, collecting primary data often include the method or procedure followed and any approximation used so that one can find its limitations. On the contrary, secondary data usually lack such information.
Despite these advantages of primary data, secondary data are extensively used particularly when a large number of items are required.
The secondary data seems to be of minor importance, specially when collection of primary data is much expensive and time-consuming.
Secondary data invariably give less meaning of the statistics and frequently present no explanation other than the captions and footnotes in the tables. In fact, some information are usually suppressed in secondary data.
Advantages of Secondary Data:
Practically, there are various advantages in using secondary data. First, cost of collection of data is less. That is why data produced by the Government, companies and various organisations are readily available. Secondly, one obtains a great variety of data on a wide range of subjects.
Thirdly, much of the secondary data available has been collected for many years and, therefore, it can be used to study the trends. Fourthly and most importantly, secondary data is of great value to the government, business world and industry and also for research organisations. Again, secondary data help the government in making present policy decisions and also planning for future economic policies.
Limitations of Secondary Data:
In spite of all these advantages of secondary data, one must use it with enough caution. First, the method employed for the collection of such data is often unsatisfactory. That is why, secondary data in most cases, is subject to transcribing errors (i.e., errors occurring due to wrong transcriptions of the primary data). Secondly, secondary data are really mere estimates and not the facts.
In view of this, secondary data suffer from estimating errors. Thirdly, scrutiny of the secondary data is obviously essential since errors may creep in due to unwanted bias. This is because of the fact that often fictitious figures are recorded unknowingly in secondary data. In this sense, secondary data are not only inaccurate but also incomplete and inadequate. Without detailed scrutiny of the secondary data one must not be advised to use them.