Primary research refers to research that is done from scratch. In other words, researchers are in the field gathering information for the first time --- raw data. Secondary research consists of mining and analysing existing research in the form of reports, articles, literature and other sources to find data to support a specific goal. The two forms of research are different and pose different challenges.
Primary research is more difficult and time consuming to conduct than secondary research. It requires original experiments or field work to come up with new data. Since secondary research is based on existing data that has been collected through primary research, it isn't considered as difficult to conduct.
In addition to being more difficult to carry out, primary research is also more costly in general. Because it takes more time and uses more resources, that translates to higher financial costs. Secondary research requires only that you locate the existing data and extract the information required for analysis. It can often be performed by lower-paid, less specialised personnel than primary research, further reducing costs.
There are a few drawbacks of secondary research related to accuracy and reliability. First, margins of error or standard deviation from the original research may not be published in secondary sources. Without this information, statistics can be misleading. Additionally, the researchers in the original study may not have had a need for great accuracy, especially when the research is market based and only a certain degree of accuracy is sufficient to make a decision. Finally, it can be difficult to find recent sources on specific topics.
Reliability and Source Bias
Relying on existing research transfers errors or misleading information present there into the results of the secondary research. It is also important to evaluate sources for bias. In some situations, researchers or organisations may have an incentive to put a certain spin on research results. For example, a commercial entity may have a vested interest in compiling research in a way that specifically supports a certain viewpoint beneficial to their marketing and profitability.
Another problem with secondary research is understanding definitions in the primary research. The information gleaned from original research is only useful with an accurate definition of the terms and exactly what is encompassed by each statistic or piece of data.