Secondary analysis is any analysis that the primary analyst doesn't do him/herself. It can be statistics, theories or hypotheses done by previous or current researchers.
Pros: Secondary analysis can help support the original analysis with additional surveys, statistics, polls, and so on. Science is based on this idea. A hypothesis gains support and credit when it can be tested and supported by other (secondary) scientists. Current science relies on previously recorded ideas and statistics. Previously recorded ideas and statistics are also examples of secondary analysis. Scientists are always aware that previously recorded data may be incorrect. For example, back in the day, the consensus was that the world was flat. But using data that’s already been tested and recorded saves the scientist time.
Cons: The problem with secondary analysis is primarily one of trust and validation. By definition, the primary analyst can do background checks and research, but must ultimately accept that the secondary analysis is true and/or done with objectivity. Also, the primary analyst must be aware of agendas or biases that might be influencing the results of any secondary analyses they decide to use. If a study was influenced by some bias, ideological aim or an agenda, this can invalidate the results. It conflicts with objectivity or it may conflict with an agenda that the primary analyst has.
Pro: If the primary analyst has an agenda, using relatively objective secondary analysis can be beneficial since it might cause the primary analyst to recognize his/her own biases.