Primary Data Vs. Secondary Data

1. Compare primary data and secondary data.

2. Discuss the advantages and disadvantages for each.

Reference:

Advantages and Disadvantages of Secondary Data

The most significant advantages of secondary data are the time and money they can save. If the information being sought is available as secondary data, the researcher can simply retrieve the data from internal or external sources and proceed to data analysis and/or interpretation. Relative to collecting primary data on the topic, this should take much less time and involve little cost. With secondary data, the original compiler of the information has already paid the expenses incurred in collecting the data. Even if there is a charge for using the data (unlike statistics compiled by government or some trade associations, commercial data are not free), the cost is usually much less than if the firm collected the information itself.

Secondary data are not without potential problems, however. Two problems that commonly arise with secondary data are (1) they do not completely fit the problem and (2) they are not totally accurate. Because secondary data were collected for other purposes, they may not fit the problem as defined. In some cases, the fit will be so poor that the data are completely inappropriate. Usually, poor fit is due to one or more of the following problems: different units of measurement, different class definitions, or the age of the data.

The size of a retail store, for instance, can be expressed in terms of gross sales, profits, square feet, and number of employees. Consumer income can be expressed by individual, family, household, and spending unit. If you need individual-level income but only household income is available, it may be necessary to collect primary data to get the information you need. Even if the units are consistent, you may find the class boundaries presented are often different from those you need. For example, if you need individual income measured in fairly small increments (say, 0 to $4,999, $5,000 to $9,999), it wont help much if the available secondary data provides individual income using wider boundaries (0 to $19,999, $20,000 to $39,999, etc.). Its also possible that secondary data are out of date. The time from data collection to data dissemination is often long. For example, government census data have great value while current, but this value decreases rapidly with time. Remember, the broad U.S. Census of the population is conducted only once every 10 years.

The other common problem with secondary data is that it is sometimes difficult (or impossible) to judge their accuracy. As youll learn throughout this book, there are lots of sources of error possible in the collection, analysis, and presentation of marketing information. When a researcher is collecting primary data, firsthand knowledge helps in gauging the accuracy of the information being collected. But when using secondary data, the task of assessing accuracy is more difficult. It helps to consider the source of the data, the sponsor of the research, and the general quality of the data collection methods and presentation.

Secondary data can be secured from either a primary source or a secondary source. A primary source is the source that originated the data. A secondary source is a source that in turn took the data from a primary source. When possible, always use the primary source of secondary data. Take this recent headline as an example, Were Seriously Underestimating the Virtual-Reality Market. The headline alone would suggest that, at a minimum, virtual reality (VR) should be the focus of the articles discussion. Upon reading the first paragraph you learn that VR is not the sole focus. Instead, its a combination of virtual and augmented reality (AR). You also learn that AR and VR combined is expected to be a $150 billion industry in five years. Consulting the primary source and author of the original study, you learn that AR is forecasted to be worth $120 billion, leaving VR to be worth only $30 billion. Technically, the secondary source headline is not incorrect. We could still be underestimating the VR market; however, since AR is expected to be four times larger than VR, doesnt the primary source provide a more interesting discovery?

A second way to judge the accuracy of secondary data is to pay attention to which organization sponsored the research. Consider the following example: According to the results of a survey, almost one-third of the respondents have talked to a doctor about a treatment they saw advertised, and about 75% of the respondents said ads for prescription drugs showed both the risks and benefits of the medicine. The results were released at a time when the U.S. Food and Drug Administration was preparing new guidelines for advertising prescription drugs directly to consumers. The survey was sponsored by Prevention magazine, which makes money in part by selling advertising to prescription drug companies. Does the presence of a business interest mean that the research doesnt accurately reflect consumer attitudes and behaviors? No, but it does mean that you need to look very closely at how the data were collected. Research that has been collected in such a way that the results will support a particular position is often referred to as advocacy research. Again, not all data collected or sponsored by an interested party should automatically be rejected as advocacy research. But pay attention.

A third way to gauge the accuracy of secondary data is to look for evidence that the research was done properly. For example, a user needs to understand how the data were collected. A primary source should provide a detailed description of the data collection process, including definitions, data collection forms, method of sampling, and so forth. If it doesnt, be careful! Such omissions may indicate sloppy methods (at best) or advocacy research (at worst). When the details are provided, examine them thoroughly. Youll need to be familiar with the research process and the potential sources of error in order to gauge the quality of secondary data. The remainder of this book provides much of the needed insight for evaluating secondary data.