In order to assess whether statistics are needed to understand data, we need to ask ourselves to what extent can statistics be useful and what happens if we don’t use statistics at all. So first of all, what is data? According to the free dictionary (http://www.thefreedictionary.com/data), data is “factual information, especially information organized for analysis or used to reason or make decisions”. Statistics is the tool which helps turn data into more intelligible information. It can therefore be useful to understand data, but is it necessary?
Raw data often contains more information than needed. It is indispensable to distinguish the useless data from the appropriate data for our findings. The selected data then needs to be analysed, organized and presented correctly before it can be fully understood. This is part of the process of understanding the meaning of the result to a study. This is true not only of scientific studies but of surveys made in any other field (economical, political, industrial, social).
Thanks to statistics, information (so data) can be processed and presented as facts. Here is an example:
The statistics pretty much speak for themselves here. If the raw data were presented to the general public, then it would not have been so easy to understand. Statistics facilitates the understanding and helps to draw inferences from data. When raw data is complex, then “data mining” can be undertaken. The term “data mining” refers to the use of statistics in order to find general trends and patterns. This is a procedure a lot of companies would go through in order to check the outcome of an event. For instance, what is the likelihood of Nivea consumers buying their usual handcream knowing the price has gone up by 16%. It thereby consists in the transformation of data into information, which is valid, could not necessarily be predicted with raw data only, and is easily understood by others. This not only facilitates the understanding but also the learning of data (such as the estimation of how many people will turn up to the lecture on Monday) or understanding the relationship between probability and statistic.
Testing the hypothesis and null hypothesis is done statistically (through a significance test). In psychology for example, 5 year-olds run faster than 7 year-olds. This can also be applied to other fields such as sociology or economy: in testing whether women spend significantly more time doing housework and less time at work than men. The statistical importance will help understand if the hypothesis was valid or not. To further stress the importance of statistics, here is a quote by H. G. Wells: “Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write”. (from a very interesting article : http://www.tdan.com/view-articles/5226, paraphrased from Samuel S. Wilks’ words (1951) in Mankind in the Making (1904, p. 192) )
However, statistics are not sufficient for understanding data as numbers do not explain reasons for the occurrences of events or facts. Statistics therefore provide a quantative but not a qualitative understanding of data. What’s more, statistics are not always necessary for understanding data when it is very simple and straight forward. In fact, data is part of every day life and numbers don’t always make an appearance in these situations. The number of eggs I ate this morning, the colour of my favourite t-shirt, or the title to the song I last listened to: all of these examples are pieces of information and therefore data. I did not need statistics in order to collect or understand that information.
In conclusion, while statistics is not in all cases necessary,it may be needed to understand your data so it is important to know how to use and interpret statistics.