What is the meaning of this lowercase "p"?

Most health professionals are resistant to statistics. There is nothing inconsistent about it: if they liked Mathematics as a profession, they would have chosen a career in the exact sciences. Clinical dentists are not different: most of them prefer to avoid greater proximity to, and as a consequence , understanding of statistics. Undoubtedly, many statistical procedures involve complex and difficult to understand calculation for those who do not have expertise in this area. It turns out that, similarly to the fact that using a car does not require knowledge of the car's intricate mechanism, electrical and electronic components, using statistics to interpret the results of a research does not require knowledge about details of the formulas that have led to these results. An important concept in the interpretation of an article is to distinguish " descriptive statistics " and " inferential statistics. " Descriptive statistics are all procedures that aim to describe the data collected in the sample. The main statistical resources in this area are: tables, graphs, parameters such as the mean, standard deviation, correlation coefficient, absolute frequency (n), and relative frequency (%). Inferential statistics appear in the article as the p-value (probability), helping us draw conclusions from survey results obtained from the population represented by the sample. As an example, suppose the researcher wants to investigate the effect of certain orthodontic treatment on the value of patients' SNA angle. When an article says that the pre-treatment mean SNA of a sample of 30 subjects is equal to 81.5°, this is the average value found by the researcher in the study sample. Assuming that the post-treatment mean is 80.5°, the researcher can conclude that, among the patients treated during the study, there was a mean reduction of 1° in the SNA. These results are descriptive statistical procedures. Descriptive statistics describe the data collected and, therefore, refer only to subjects in the research. This description does not involve any calculation of probability, since proba-bilistic error is not involved at all. Thus, it can be concluded that the treated patients had a mean reduction of 1°, because the researcher has observed this. In a survey, however, we want to reach a conclusion not only for those who participated in the survey (sample), but also for all other subjects who did not participate in the study (population) but may need the kind of treatment studied. To make a statistical inference is to …