The Differences Between Precision, Accuracy, Reliability, and Validity

In general, there are a lot of confusions in the medical and scientific world among some of its technical words. After data processing, measurement processes usually involve the terms precision, accuracy, validity, reliability, and variability. This part of the paper aims to properly define each one while citing examples in the field of OHS (Occupational Health and Safety).

Sometimes, a study has to be repeated a number of times in similar circumstances. This creates some degree of resemblance of the results. It is called precision. This is the preferred word in modern science or medicine in terms of the extent to which information that is the same is elicited in a measurement and is constantly repeated. Precision is usually mistaken for other things used to describe a result in various methodologies. The most common one is the term accuracy. While that confusion would be described later in this paper, it should be noted that the main difference is the way that measurements are compared to each other. Precision is related to how close each one is to other results. On the other hand, accuracy would be a gauge on how close such data are to the true value. Precision is also confused with reliability. The main difference lies in how results are taken and measured. It should be noted that while some data are reliable, it can be imprecise (Streiner & Norman, 2006). At the same time, a result may be precise but unreliable (Streiner & Norman, 2006). Other terms are also confused with precision. An outdated version of validity is usually regarded as synonymous to being precise (Streiner & Norman, 2006).  

Some people in the field of writing medical or scientific journals mistake the terms “validity” and “reliability” with “accuracy” and “precision” (Streiner & Norman, 2006). It is important to make a distinction among those words. As mentioned above, accuracy refers to how close a measurement is to a specific value and precision simply means how similar each result is to the others. On the other hand, validity can be described as the extent in which calculations are investigated according to its intended function. Simply put, it means that results should refer to the supposed measure of a data. In hindsight, they could be considered valid if they are supported by a specific expected outcome. Reliability means that the same values should appear even if the process of measurement is repeated. Data collected are considered reliable if each time a calculation or analysis is performed, they yield similar results. 
Validity can be generally referred to as an adjective to describe something as justifiable and grounded (Swanson, 2014). Some people believe that this term may be similar to accuracy in terms of statistical data (Swanson, 2014). Validity can be considered as not having any bias. Thus, it should be used to describe a measurement that is supposed to be measured according to an intended universal truth. There are actually two kinds of validity. The first one is internal in which the question of whether the results are valid for specific subjects are asked (Swanson, 2014). On the other hand, there is the other kind called external validity. It is synonymous to generalizability. The population of a specific sample is used as the comparative basis for the validity of a result.

Reliability refers to the consistency in which results or data are gathered (Swanson, 2014).  As mentioned above, it refers to the ability of a measurement to be repeated when the same exact procedure is performed. One should get the same values if calculations are done repeatedly.

While a lot of literature describes reliability, validity, accuracy, and precision, there is little description written for variability. In hindsight, it means the ability of a measurement to be varied in both similar and different processes (Dale, Rohn, Patton, Standeven, & Evanoff, 2011).

As applied in the context of OHS, accuracy can be applied to information regarding determining whether a specific injury can be directly related to an event. In the area of construction, one may consider if head injuries are accurate to inconsistent accident measures of a location. One may ask whether safety nets are properly placed. 

Precision can be applied to how a specific sickness may be affecting workers in a factory. This may include the fact that everyone of them has their lungs affected and has similar fever temperature. It may not be necessarily accurate to the location of their labor. Although it is precise, it may come from various sources or it may passed from the virus of one person to another. 

Validity can be applied to an office scenario. Each employee may have reasons for not attending work on the same day. The manager just needs to find out if they are valid. Thus, the values presented by the causes of the illnesses of each must be similar to what is supposed to be. If their reasons are all related to a company outing they attended yesterday, it should be valid. If they got flu or cold because they swam in the pool, it may be describing what was supposed to be measured.

Reliability of information in OHS is another tricky concept. It is especially applicable to those who suffer from temporary disability due to some accident. Their work can be evaluated on both the time they were disabled and not. Through repeatedly getting the same data from a specific measurement of productivity, the reliability of the worker can be determined. 

Aside from reliability, variability can also be used for employees who have temporary disability. One could gauge if their work production would have any variance before and after the injury or accident. Such data can be used to see if the employee can still work despite variability in his/her productivity. It is also critical to see if there is variance in output. Another example is a night shift factory worker who can produce a specific number of units. Variability can be measured in different times, especially during midnight where everyone might feel sleepy.

The Scientific Method

Observation

The observation step in the scientific method is one of the essentials in this process (Baumgardner, 2008). While modifications for this procedure had been developed nowadays, this concept remains an integral part of today's related analysis. This is the foundation of such an investigation and would be applied accordingly in the case of workers complaining of breathing problems in the mining site. 

The main observation can be summarized by the facts already presented in the case. There may be additional information needed but the provided is the only data that this paper can work with. The complaints happen by mid-morning when breathing difficulties are experienced by mining workers inside the site. It happens specifically during the end of a working week. Furthermore, absenteeism rate in this group is considered higher than those who have worked above the mining grounds. It is important to consider such an observation in the light of the condition of the mines. The air in the area should be tested and the reaction of people to it should be observed.

Asking a Critical Question

The critical part of the scientific method would start with the asking of a critical question (Li & Gruenert, 2014). In the case of the workers' complaints in the mine, it is important to ask a question that corresponds to the observation mentioned above. This would basically start the scientific inquiry. It is important that it is not too broad in order not to include irrelevant things. Being too specific may not be good as well since it may not answer the question directly as needed. Rather, it might provide an answer for some technical things related to it. 

The critical question that needs to be asked is if the complaints of the workers about breathing difficulties towards the end of the work week and the high rate of absenteeism in that span of time inside the mines is related to a specific condition of air inside their environment or if it is just an excuse made up by the majority. This inquiry should tackle precision and validity according to worker's complaints and if there are real grounds for it.

Developing a Hypothesis

A hypothesis should be mutually exclusive to parties involved in order to come up with a legitimate analysis (Rosen, 2016). In simple terms, this can be described as an intelligent guess that needs to be proven. In the case of the mines, it could actually be a hypothetical answer to the question asked above. 

While it looks like the workers have real reasons to complain about breathing difficulties, the timing and the rate of absenteeism seem to imply that they are just making excuses not to go to work at specific times. This hypothesis statement would be tested and studied in order to prove the critical question and possibly develop a theory. Its development came from the fact that the timing of such complaints is a little bit suspicious. While OHS should consider the health implications of such an incident, it is important to consider why the complaints only come during the end of work week and how come the rate of absenteeism is higher to those inside the mines than outside. The condition of the air within the working environment should be studied, as well as the motivation of the laborers. It is important to discern if they are really experiencing that and why the situation only happens during the end of the work week.

Making a Prediction for Testing

A prediction needs to be consistent with the concepts imbibed in the hypothesis (Ayala, 2009). For the case of the mines, it would need to identify the validity of the breathing difficulty complaints. It should be noted that it creates higher absenteeism than those above ground and that it only happens during end of the work week. As a prediction, the breathing difficulties might actually be caused by the condition of the air within the mines, but the complaints are only purposely done at the end of the work week and are being used as an excuse for being absent. It is not an unusual thing for workers inside an enclosed environment to experience such problem. The breathing difficulties may be a valid reason. However, the timing is a little bit suspicious and unreliable. It could be noted that it may be an excuse for workers to leave work early. In the first place, difficulty in breathing is expected inside a mine. However, the principles of OHS still require a thorough examination of the air condition inside the enclosed structure. This is to make sure that the environment is safe for laborers. 

Experiments to Test PredictionTo test prediction, the work week pattern can be changed. This is critical in determining the accuracy of the breathing difficulties only happening by the end of such time. As an experiment, the work week could be altered through changing the start and end date. If for a specific time the miners would work from Monday to Friday, they can change that. On the coming week, it would be from Tuesday to Saturday and so on and so forth. This experiment would allow investigators to properly gauge if breathing difficulties complaints only happen during a specific day. Aside from that, shifts could also be applied to test the prediction. They can let miners work from Monday to Wednesday and the rest Thursday and Friday before working on a weekend from Saturday to Sunday. This pattern can be mixed up. The purpose of these shifts is to see if workers would still complain about difficulties in breathing during the said time. Aside from that, it is important to test the quality of air in the working environment as an OHS requirement. It should be noted that the best version of the condition should be applied. That means that a lot of workers should be doing labor inside the mines during those times. It should be done randomly on different days in order to confirm if complaints are valid.

Data Collection and Analysis

For data collection, investigators should get the number of breathing difficulties complaints during a specific time of the day and date. They should compare it to various work time patterns described above. Data from air condition experiments should also be considered during the same duration of labor. This would then be cross-analyzed with each other. 

Initial analysis should indicate that breathing difficulties might be a normal occurrence in mines since it is an enclosed area. While conditions of air may vary, it is constantly more difficult to breath in such environment rather than above ground. The breathing difficulties complaints might be true but it is not a serious occurrence and the timing is really suspicious. Altering dates of work should show that it is just an excuse by workers to have an early weekend getaway.

Logical Conclusions Based on Experimental Results

Based on expected experimental results, it could be concluded that the workers are just making an excuse to leave work early and are taking advantage of the situation in the enclosed area. While it is not yet clear whether all the miners have agreed on this, the suspicious timing should be a giveaway. However, OHS still requires testing the condition of air inside the mines. Data in this research should be cross-analyzed. This analysis is needed for further study. However, the change in work day patterns should give more than enough information regarding the validity and reliability of the incident.

References

Ayala, F.J. (2009). Darwin and the scientific method. Proceedings of the National Academy of Sciences of the United States of America 106(Suppl 1), 10033- 10039. 

Baumgardner, J. (2008). Exploring the limitations of the scientific method. Arts & Facts 37(3),  4. 

Dale, A.M., Rohn, A.E., Patton, A., Standeven, J., & Evanoff, B.A. (2011). Variability and misclassification of worker estimated hand force. Applied Ergonomics 42(6), 846- 851. 

Lake, L.W., & Bryant, S.L. (2006). The scientific method and earth sciences. Journal of Energy Resources Technology 128(4), 245- 246. 

Li, L.,  Gruenert, D.C. (2014). Scientific methods- the foundation of science. Journal of Biological Methods 1(1), e4. doi: 10.14440/jbm.2014.20

Rosen, J. (2016). Research protocols: a forest of hypotheses. Nature 536, 239- 241. doi:10.1038/nj7615-239a

Streiner, D.L., & Norman, G.R. (2006). “Precision” and “accuracy”: two terms that are neither. Journal of Clinical Epidemiology 59(4): 327- 330.

Swanson, E. (2014). Validity, reliability, and the questionable role of pyschometrics in plastic  surgery. Plastic and Reconstructive Surgery- Global Open 2(6). e161.