Fig.1 The answers in one poll concerning Global Warming 3.2. Intermediary stages: data organization, analysis and presentation Once a study moves from the initial phases of planning, design and data collection, scientists must then organize the information in a rational, relevant and logical way. One area of concern is the way in which data is stored and the nature of the study: ideally, special attention must be paid to storing the data in a way that does not allow an individual to be associated with a potentially vulnerable group . This prevents potential misuse of data that may ultimately result in harm to individuals; among the examples most frequently cited in scientific literature regarding the extent of potential harm are those concerning abuses in the field of medicine, which resulted in extraordinary harm to political or social groups at risk (such is the case of the medical experiments conducted by the Nazis during World War II). Therefore, it is the moral obligation of the statistician to take all the precautions necessary to prevent the misuse of data that may potentially harm individuals. While the possibility of such harm occurring in present times seems distant and unlikely because of national and international laws, it is important to keep in mind that the legality of an action does not necessarily imply that an action is also ethical; furthermore, history offers important lessons on how we may prevent potentially tragic events from occurring in the future. We focus next on the use of averages and how statisticians must pay close attention to the type of averages used and their relevance in the context of the particular study conducted. For instance, let’s assume we are trying to calculate the values of land properties in a neighborhood of 100 land plots, where 90 of them are valued between 180,000 to 200,000, and 10 of the properties have values between 500,000 and 550,000. If we were to compute the mean property value, the figure would be skewed by the extremely high values of the 10 land plots, and would present an unrealistic assessment of the property values. By using the median values instead, we would obtain a result likely between 180,000 to 200,000; this offers a more realistic view, without having to remove the extremely high values (suppressing data, as we have previously mentioned, is unethical and should be avoided). Another area of particular concern is the use of graphical representation of statistical results, as graphs and charts have widely been used to misrepresent data – from political polls, to economic trends - no area of research is immune to such practices. Let us consider the two graphs below (Fig.2 and 3): they both depict a decrease in the average house values from $63,000 in 2009, to $59,000 in 2011 (note: these are fictional values used only to illustrate an example). By looking at the graph from Fig.2, it would appear that in 2011 the average house value declined to almost half of the average 2009 value. However, the numbers prove this conclusion to be incorrect.