Question

1. Reflect on the misuses of Data. How do people use biased samples and surveys that would affect conclusions? Give an example.

2. How do people manipulate data to display the conclusions? Give an example.

Answer 1

1. Data misuse is the use of information in a way it wasn't intended to be used. In today's world, data misuse is very common due to the advent of data analytics and artificial intelligence. People misuse data to track people and manipulate them in a way for their own advantage. For example, recently Facebook was in the news of tracking its user's data and information and using them to suggest advertisements and also interfering in the election results of the USA. In 2007 Colgate was ordered to abandon their claim of "More than 80% dentists recommend Colgate" as the claims were false and data manipulation was done to represent desired results. Misuse of data convinces people to believe false claims and disrupts the balance and equality and could also create huge problems due to false conclusions. A biased sample is a sample where the data collected are not random or some data collected in the sample have a higher probability than others. For example, a media house showing statistics in favour of a certain political party where they have sampled majority of the intended party supporters and their views. Thus a definite bias is created that leads to a misleading conclusion that the intended party is way ahead of the other parties based on a random sample.

2. Data can be manipulated in a no. of ways like collection, analysing and even stating the conclusion. There are a no. of ways like sorting,change of scale, immersion, averaging, etc. One simple data manipulation trick which is very common is putting in random data values of your own choice in the sampled data set to get the desired statistic. For example, in a quality check procedure where it has been stated that the no. of defectives must be under a certain limit else, it can result in the termination of the quality check administrator. In fear of losing his job, the administrator might pass several defective items as non-defective and would eventually maintain the defective no. Another way of displaying manipulated data to achieve desired conclusion are graphs, where certain data points might not be included for example some media houses shwoing their TRP ratings where they only show the upward trend of few months and not the downward trend so as to maintain a superior image infront of the viewers, or manipulating the y-axis of the graph where the scale is disproportionate to the data so as to represent a more significant or less significant than they actually are.