# 32. Statistical manipulation ## 32.1. Definition This is a group of manipulative methods used to convey incorrect or misleading messages to victims, using obscure and often very complex arguments. The arguments used and messages being delivered derive from various methodologies, like statistical or other scientific or mathematical analyses. Most of these are incomprehensible to the average victim, and this makes any kind of defence very difficult or even impossible. ## 32.2. Origins and motivation It is interesting to see how groups and individual manipulators tend to come up with a hypothesis and then work to prove it right, instead of working to prove it wrong. Once satisfied, the manipulator stops searching. This works for the individual as well. We often seek safe havens for our ideology with friends and co-workers who are like minded and have similar attitudes. We use only those media outlets which are guaranteed to reinforce our existing views. We rarely deliberately allow or encourage our conclusions to be tested, or our preconceptions to be challenged. Statistical manipulation is a mathematical and logical adjunct to this strategy of seeking confirmation because it relies on most people not being able to challenge its conclusions and therefore not bothering to even try. If the manipulator's message fits the victim's philosophical expectations, then why bother to challenge it or understand the terribly complex statistical methodology? ## 32.3. Methodology The methodologies used in statistical manipulation are an inscrutable cover to the manipulator because they are effectively opaque to the vast majority of the victim population. Because of this, a victim is obliged to accept a seemingly credible message simply because they have no way of verifying or denying it. These types of manipulation are considered "complex" because they employ carefully chosen methods which require that the manipulator has an advanced understanding of certain scientific methodologies, such as statistical analysis. They rely on the generally poor understanding of these subjects by the victim (often the public) and the fact that the cost to the victim of disproving an assertion is too high and maybe just too difficult. These manipulative methods also rely on the fact that the underlying manipulative scandal, if it is ever revealed, is often too complex to communicate in a single sound bite. Thus, this type of manipulative act doesn't always get the instant attention or interest of the victims that it deserves. These techniques are effective in obscuring a proper intellectual understanding by the victim of some central reality of the chosen subject. All scientific methodologies can be used to cover some attempted manipulation. Science is now so specialised and generally so poorly understood by the general public, that a manipulator can deliver an important set of messages before anyone gets time to contradict them. These days we are constantly bombarded by seemingly credible study results about our health and our diet. Many of these studies are complex, many have seemingly contradictory results and all claim a certain degree of accuracy. However, we don't totally understand the study, the results or the degree of reliability. For this reason, none of the sciences are better suited to the purpose of widespread manipulation than the use of statistical analysis. This is a subject so open to debate and so poorly understood that it is ripe for misuse by a manipulator. And so we have focused here on the misuse of statistics. The use of complex manipulative techniques should not be confused with the manipulation derived from the "politicisation of science", which is related but different. We deal with that subject in a separate chapter. ## 32.4. Methodology/Refinements/Sub-species *See Child Pages:* - [[Biased Samples]] - [[Biased Sample Selection]] - [[Fox Guards the Chickens]] - [[Lies of Omission]] - [[Selecting Favourable Criteria]] - [[Self-Selection]] - [[Data Manipulation]] - [[Cherry-picking Results]] - [[Creating or Altering Data]] - [[Losing Data]] - [[Discard Unfavourable Data]] - [[Dump Data]] - [[File Drawer Effect]] - [[Positive Outcome Publication Bias]] - [[Switching Dependent Variables]] - [[Trial Termination]] - [[False Causality]] - [[Clustering]] - [[Regression Fallacy]] - [[Data Dredging]] - [[Gamblers Fallacy]] - [[Loaded Questions]] - [[Misreport Error]] - [[Non-enduring Class Fallacies]] - [[Over Generalisation]] - [[Proof Null Hypothesis]] - [[Size Matters]] ## 32.5. Avoidance and Counteraction The subject matter and methodologies used to deliver statistically manipulated messages is so difficult to understand for most of the general public that this form of manipulation is very hard to detect, avoid and contradict. Again, this method relies on the very high cost to the victim of disproving the manipulator's assertions. There are two possible avoidance and counteraction strategies for the victim: - The first is to become personally familiar with statistical analysis and methodology. This may be a tall order for the average person. - The second is to enlist a professional statistician to undermine the manipulator's arguments, if they are based on illicit statistical assumptions. Fortunately for the victim, a properly phrased and credible denial of a statistical manipulator's assertions can deliver a knockout blow to the manipulator at a relatively low cost to the victim. A victim needs to deliver this counterattack using highly credible and respectable expert sources in a properly argued case. A victim also needs to make their response more understandable and more reasonable than the manipulator's original assertion.