The morality of statistics: Will statistics invariably lie?
A famous book from the 1950s was entitled How to lie with statistics. Implied by the counter-intuitive name was the concept that the old cliche that 'numbers don't lie' was false. In fact, as discussed in the article "Reflection before action: The statistical consultant confronts ethical issues" by S. Andrew Ostapski and Claude R. Superville, statistics can be highly subjective in terms of how they are presented as are the conclusions which can be drawn from them. Even researchers have been accused of manipulating statistics to prove 'facts' that are not true within academia. The pressures only increase when statisticians are asked to serve the financial 'masters' of commerce. "The ability to be creative in building interdisciplinary bridges can be risky, especially when the parties that are served do not understand the statistical process. The statistician must not only make sense out of the data but also develop the means to insure the proper interpretation of such information by all relevant parties" (Ostapski & Superville 2001).
For example, a statistician may be solicited by a company wishing to prove the effectiveness of a new workout program, demonstrating that statistically significant results occur when persons use the new exercise machine developed by the company. However, there may be many problems with the designed study, namely the fact that the experimental group in the study was placed on a strict diet while the control group was not. Also, the fact that the group using the workout machine experienced better results than the control group might be attributed to the fact that some form of workout is better than no workout, but this does not mean that people need to buy the expensive contraption to have the same results. Finally, there are issues in experimental research such as the fact that the mere act of being observed may cause persons to experience effects (i.e., to change their behaviors). Without taking note of these factors, simply saying that the program was effective does not paint a full portrait of the reality of the results. Statisticians do not have a code of professional ethics like professionals in law and medicine, so there is a certain degree of subjectivity about any decisions they make. However, the impact of statistics can have far-reaching consequences if deployed in an unethical manner. A study which 'fudges' the results of the potential side effects and dangers of a drug could result in the loss of lives of millions; even public policy decisions are made based upon the implications of statistical data.
A Kantian or deontological fashion of approaching questions related to ethics suggests that the person must ask if there is a duty to act and would the person be willing to act in the chosen manner as if setting a universal law for all time. In contrast, a utilitarian approach would take a cost-benefit analysis in terms of agreeing to analyze something using a particular statistical approach -- do the benefits of doing so outweigh the harms, asks the consequentialist? Still other methods of assessing ethical behaviors include that of community standards (would "the decision be the same if it were published on the front page of The Wall Street Journal?") or if such actions fit the Golden Rule (would they be acceptable if 'done unto me;') (Ostapski & Superville 2001). The varied use of such different rubrics of analysis is likely to produce very different ethical decisions.
If one takes a deontological view that all misrepresentation is wrong, than any statistical analysis that does not paint a full and forthright picture of the matter being studied is immoral. The much-touted statistic that '9 out of 10 dentists recommend' a particular brand of toothpaste is 'wrong,' even though the consequences are arguably relatively minor and constitute accepted 'puffery' for an advertisement. Even statistics that could potentially motivate people to act in a questionable manner might be considered unethical, if they are likely to be misread, despite the fact they are technically correct. For example, during flu season, it is widely regarded as beneficial to the community if as many people who can get vaccinated do so (provided they are not allergic to the vaccine). However, there will always be some reported side effects which may or may not be attributable to the effects of the vaccine. A deontological view would suggest that a statistician has the responsibility to report the truth, even if this may discourage people from getting the vaccine: there is a psychological tendency to over-value negative effects of action vs. inaction. A 'Golden Rule' analysis might similarly support such honesty, since most of us 'want to know' if there is a potentially bad consequence of something, however slight the possibility. In contrast, a utilitarian view would suggest that the potentially bad effects of many people not getting vaccine are so great, the statistical results on the side effects should not be reported, or be reported in 'fine print' in a dismissive fashion, to ensure people are not unduly alarmed. However, a utilitarian might have difficulty justifying dismissing statistics about the side effects of a new drug being produced by a business just so the business could encourage people to use the drug to make a profit, even though the side effects were potentially serious and taking the drug did not in actuality produce benefits for the patient comparable to treatments with fewer side effects.
Even when technically partisan, paid, or supported by an institutional authority, the statistician must still remember that he or she has a 'higher calling' to his or her own personal ethics, to the ethics of the profession, and a responsibility to humanity. "The client may have a pre-conceived notion of which results a particular study should conclude...it is ethical for statisticians to report all results, both those in favor and against his client's views, they may have no control on how the results are used and whether selective reporting of the results will occur" (Ostapski & Superville 2001). Even if not technically 'in control' the statistician can do all he or she can do ensure that the most complete presentation of the data is achieved and avoid presenting the data in a manner that can be easily skewed.
Given the potential for misuse of statistics, it might be tempted to argue that statisticians should only use their methods to compile completely objective data and keep analysis to a minimum. However, interpreting statistics is a learned skill and not every member of the public has the ability to assess raw data. Without statistical analysis, a very important component of human knowledge would be lost and thus statistics must be presented in a comprehensible form. Thus, although statistics can 'lie,' the Christian theologian Jan Geertsema notes that they can offer many potential moral uses, particularly in a democracy in which individual, free will is a cornerstone of the worldview of the society. "A Christian point-of-view is very relevant in this controversy. Opinion polls can be of great help to a government that is committed to the biblical norm that those who are in a position of authority have the duty to serve those over whom they have authority" (Geertsema 1987)
In a business context, it could be argued that people must be able to rationally evaluate the evidence presented to them by a business that a particular product, service, or treatment is useful or not useful. This is why 'puffery' in advertisements to some degree is allowed -- advertising a sandwich as the 'world's best grilled cheese according to 100% of our satisfied customers' is not regarded as immoral, because within the context it is assumed to be a normal part of advertising and the intelligence of the individual…