What is Data Analysis?

What is data analysis exactly? Try to define the phrase, data analysis and you’ll get different answers depending on the discipline or industry of the person you are asking.

A scientist or sociologist may define it as a body of methods used to help describe facts (based on observation), detect patterns, develop explanations or test a hypothesis.

A marketing manager may define it more closely to the Wikipedia article, “Analysis of data is a process of inspecting, cleaning, transforming, and modeling data with the goal of highlighting useful information, suggesting conclusions, and supporting decision making.”

Ask a mathematician or a statistician and they may well tell you that to define data analysis one must first define statistics, which, may be defined as a set of methods used to collect, analyze present and interpret data.

So again we ask the question, “What is Data Analysis?”  We define it as the process of taking raw information (data) and creating something understandable and useful, that either supports or rejects a specified goal.  In other words, when the CEO asks, “Should I or can I do what I want to do?”; analysis of the available data will support either a yes or no answer.

Analyzing Data For The Real World

Although raw data, expressed as a numerical result, describes values that relate to questions, in the real world the analysis of this data is not about the numbers.  Rather it is about translating those numbers into something meaningful that relates to a real world question.

Your analysis may indicate trends or show an average, median or mean. It can identify groups or sub groups with common characteristics or differences among those groups.  Data analysis can show the rate of growth or the rate of decline.  When translating these numbers into an answer for a real world question, a skilled analyst can present information in such a way as to significantly influence a decision maker to move in a certain direction.  However, using the same data, it is possible to make a strong case for going in the opposite direction. This apparent contradiction is not found within the data; rather it is found within the motivation, background and bias of the analyst and/or within the political and economic constraints of the analyst’s organization or client.

Analysis Presents a Position

Ultimately, data analysis is used to present an argument.  Should the company promote product A?  Should it raise prices?  Can we sell more of product A if we introduce change B? These types of ‘Should we’, ‘Can we’ questions require the support of data that has been collected, analyzed and presented in as unbiased a manner as possible.  The data that is collected is rarely, if ever, a Should we, Can we question.  Most often the Should we, Can we questions are answered in response to data collected relating to Who, What, When, Where or How questions. (Who buys hair spray?  What is the median income of our clients? Where do they shop? When do they shop? How do they shop? etc.)

Bias and Mistake

It’s up to the analyst to interpret and present the data from the Who, What, When, Where and How questions to the decision maker in such a way as for the decision maker to answer the Should we, Can we questions.  For the reasons stated above, presenting a wholly unbiased interpretation is nearly impossible. Often it is not the process of analysis that is at fault but rather the conclusions drawn from the information (or how they are presented) that lead to the failure of a product launch or ad campaign.  A primary culprit leading to erroneous conclusions is mistaking correlation for causation.  The classic example: roosters do not cause the sun to rise!    There are, however, numerous techniques and tools analysts use to minimize errors but those require additional explanation not suited for this post.

What is Data Analysis Then?

So we ask one more time, “What is Data Analysis?”  I would argue that data analysis is simply the interpretation and presentation of the validity or invalidity of data as it relates to a question posited by a decision maker who needs to understand and rely on the information presented by the analyst in order to take action on that question.

In other words, data analysis is the telling of the data’s story. It is a story that began with the questions, “Should we or Can we do what we want to do” and ends with the answer, “That depends on your interpretation”.

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>