In scientific research, accounting for both the social sciences and the natural sciences, there are two main types of research analysis: qualitative and quantitative. Qualitative research is used in social science to help draw conclusions about a topic and relies heavily on observation and inferences, rather than attempting to directly quantify data. Quantitative research, on the other hand, is usually relied upon in the natural sciences — and at times in the social sciences — to directly measure research results, often assigning exacts measurements. Between these two extremes is semi-quantitative analysis, which assigns approximate measurements to data, rather than an exact measurement. Often used in cases where a direct measurement is not possible, but inference is unacceptable, semi-quantitative analysis has many applications in both the natural and social sciences.
An example would be if a manufacturing plant produces five times more tires for cars than trucks a semi-quantitative analysis is used, rather than an analysis that gives an absolute value. Instead, an absolute value as defined by quantitative research methods would state the exact number of truck and car tires produced every day since the plant began production. Therefore, semi-quantitative analysis is not precision driven, but rather approximately correlated. Using such analytical methods allows researchers and scientists to apply quantification where a reliable idea of a measurement is useful, although precisions measurements are not possible. In particular, the analysis is useful in cases where quantified data might fluctuate periodically.
Historical scientific study is one such application of semi-quantitative analysis. For instance, if a meteorologist wants to ascertain the annual average temperature of a specific geographical location, he or she will gather records of temperatures going back as far as recorded history will uncover. Using that data, along with current temperature measurements, he or she will establish a range of temperatures that reflects the variance throughout the year. It is more acceptable to use a semi-quantitative analysis in this instant, rather than an attempt at an exact quantification, because temperatures fluctuate both yearly and through the year due to a variety of factors.
Geneticists as well make extensive use of semi-quantitative analysis, since exact attribution of DNA is often not possible, but instead will fall with a range of probabilities. For example, in forensic science when a DNA sample is analyzed, then compared to another DNA sample, the results are expressed in percentages of probability that there is or is not a match. While not 100 percent, such analysis does provide a near certainty or approximation that can generally be relied upon.
