{"id":7050,"date":"2026-09-21T07:45:58","date_gmt":"2026-09-21T07:45:58","guid":{"rendered":"https:\/\/hacto.umh.es\/2026\/09\/21\/how-to-interpret-p-value-in-scientific-studies\/"},"modified":"2026-09-21T08:41:30","modified_gmt":"2026-09-21T08:41:30","slug":"how-to-interpret-p-value-in-scientific-studies","status":"publish","type":"post","link":"https:\/\/hacto.umh.es\/en\/2026\/09\/21\/how-to-interpret-p-value-in-scientific-studies\/","title":{"rendered":"How to interpret p-value in scientific studies."},"content":{"rendered":"

[et_pb_section fb_built=”1″ admin_label=”section” _builder_version=”4.16″ global_colors_info=”{}”][et_pb_row admin_label=”row” _builder_version=”4.17.4″ background_size=”initial” background_position=”top_left” background_repeat=”repeat” global_colors_info=”{}”][et_pb_column type=”4_4″ _builder_version=”4.16″ custom_padding=”|||” global_colors_info=”{}” custom_padding__hover=”|||”][et_pb_text _builder_version=”4.17.4″ _module_preset=”default” global_colors_info=”{}”]P-value is at the same time one of the most used and misunderstood concepts in research. That\u2019s why in this post we will explain what it is, which information it provides and how it should be interpreted alongside the rest of the results.<\/p>\n

When working with a sample population, variability due to randomness is inevitable. Consequently, the differences or associations observed may be indicative of a characteristic of the population or they may be the results of random variability.<\/p>\n

Example: <\/strong>Suppose we wanted to find out whether taking part in an occupational prescription programme reduces loneliness in older adults. After comparing participants with non-participants, we can observe that the participants present lower levels of loneliness. However, as we have only observed a sample of the population, we cannot guarantee that this difference is in fact due to the programme. It may be due to the sample\u2019s variation.<\/p>\n

This is where the p-value, alongside other indicators, will help us to interpret the results.<\/p>\n

Previous considerations<\/u><\/p>\n

\u201cHypothesis testing\u201d is a method to find enough evidence to reject the null hypothesis (H0) and support the alternative hypothesis (HA). The HA is usually the hypothesis of interest, referring to the existence of difference, association or effect between the variables of the study.<\/p>\n\n\n\n
For example (HA):<\/strong> Falls prevention\u2019s programs reduce the number of hospital admissions amongst older adults compared to regular care.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

On the other hand, the H0 is established opposed to our research interest, indicating an absence of difference, association or effect.<\/p>\n\n\n\n
For example (H0): <\/strong>Falls prevention programs do not reduce the number of hospital admissions amongst older adults compared to regular care.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

However, it is important to point out that the scientific statistical community advocates for a more careful interpretation of this element (p-value). It should be considered alongside other indicators of the study such as the confidence intervals (CI) (Wasserstein & Lazar, 2016).<\/p>\n

 
\n[\/et_pb_text][et_pb_image src=”https:\/\/hacto.umh.es\/files\/2026\/09\/PublicaTO-Infografia-1.jpg” title_text=”PublicaTO Infografia 1″ _builder_version=”4.17.4″ _module_preset=”default” global_colors_info=”{}”][\/et_pb_image][et_pb_text _builder_version=”4.17.4″ _module_preset=”default” global_colors_info=”{}”]Imagine that an occupational therapy programme reduces the feeling of loneliness on a scale by 4 points. This estimation is based on a sample and therefore, it is subject to uncertainty. The CI indicates the range of values which it is reasonable to assume that the programme\u2019s real effect occurs in the population. The narrower the range, the higher the accuracy of the estimation. However, we should not confuse this range with the interpretation of a 95% probability that the true value lies within it.<\/p>\n

P-value\u2019s interpretation<\/u><\/p>\n

The p-value is calculated based on the chosen measure and the hypothesis test that accompanies it. In 2016, the American Statistical Association (ASA) publicised a historical declaration to clarify the use of the p-value (Wasserstein, & Lazar, 2016). They remind us that the p-value is a useful tool but it should never be interpreted alone. The ASA recommends interpreting the results of a study alongside the effect size, the CI, the methodological quality, the study design and the scientific context. Therefore, it is not enough to make conclusions based on the p-value being higher or lower than 0.05. Their declaration included 6 points:<\/p>\n

    \n
  1. P-values can indicate how incompatible the data are with a specified statistical model.<\/li>\n
  2. P-values do not measure the probability that the studied hypothesis is true, or the probability that the data were produced by random chance alone.<\/li>\n
  3. Scientific conclusions and business or policy decisions should not be based only on whether a p-value passes a specific threshold.<\/li>\n
  4. Proper inference requires full transparency and the complete presentation of data, including estimates and uncertainty, through confidence intervals.<\/li>\n
  5. A p-value, or statistical significance, does not measure the size of an effect or the importance of a result.<\/li>\n
  6. By itself, a p-value does not provide a good measure of evidence regarding a model or hypothesis.<\/li>\n<\/ol>\n

    The interpretation of the p-value depends on the alfa\u2019s significance (\u03b1) level or \u201ccut-off point<\/strong>\u201d which is usually established at 0.05 (5%). This value represents the risk that we are willing to accept incorrectly that the result of a difference or association exists, when in fact it does not. This is known as Type Error I or false positive. Based on this we can interpret the p-value as follows:<\/p>\n