The statistical methods are powerful tools for gaining information about the social or natural phenomena. However, the interpretation of the data requires careful sampling, generalization and application of the statistical test. In addition to mathematical and statistical knowledge, common sense is required.
Sampling refers to choice of respondents for the research. It is almost never possible to study the whole population. Therefore, the researcher chooses a certain group and studies its characteristics. However, if the group is chosen incorrectly, the results are inaccurate as well. For example, if the blood pressure is measured among the patients in the clinic, the researcher has to keep in mind that all age groups have to be represented equally. If most of the respondents are senior, then the results are bias.
The conclusions in public health has to be based on numerous observations since the cases of severe deceases or epiphenomena are rare. For example, tuberculosis research can be performed only basing on the national data, since there are not so many sickness cases. The research based on the regional data lacks significance. The factors that lead to epiphenomena after operations also can be studied only when the vast cases are available. When the factors are studied basing on the experience of one hospital, it is the inappropriate interpretation of the statistical data, even if the tests are applied.
The research results are often recognized as “statistically significant”, which means that p < 0.05. Nowadays, this approach is acknowledged inappropriate, since different studies require different significance levels. The scientific research states the p-value (0.001, 0.02, etc.), and allows the reader to decide which level is statistically significant (Kault, 2003).
The application of the statistical research requires careful planning, data collection, application of the statistical methods, and profound knowledge of results presentation. These factors allow the researcher to avoid the inappropriate use of statistics.
References
Kault, D. (2003). Statistics with common sense. Westport, Conn: Greenwood Press.
