This activity is important because managers should know the difference between correlation and causation. It is important to question any suggestion of causal relationships. Decision making should rely on good data and careful analysis.

The goal of this activity is to demonstrate your understanding of correlation and causation by reading a case and answering questions that follow.

Read the following case and answer the questions that follow.

Walmart found that people buy more flashlights before storms. But they also buy more Pop Tarts. Researchers say there is no causal relationship between storms and Pop Tarts, but the relationship between storms and flashlights certainly appears to be logical$.$ These conclusions don$$t keep Walmart from using strange relationships to guide its marketing, though. Weather affects sales, and the company pays attention.

Although Walmart and other retailers have been successful with some wacky correlations in the retail market, many companies don$$t pay attention to data, use bad data, or assume causation.

Some companies don$$t use evidence from past causal relationships when making new decisions. For example, Amazon famously ignored data about its successful products when producing the Amazon Fire Phone in $2014.$ Amazon typically created cheaper devices that do just what the customer needs, like the Kindle Fire, and investors assumed that the Fire Phone would be the same. The price of the Amazon Fire phone with a contract started at a pricey $($at that time$)$ $$200,$ with fancy features that customers didn$$t actually want. Lack of demand led it to drop to $99$ cents within a few months. Amazon learned that its basic devices created more sales.

Other companies make decisions when they don$$t consider all the possible variables. Google Glass, a wearable camera that looks like glasses, was a major mistake in $20132014$ due to incomplete analysis. Google assumed that a camera you can wear was cool enough that people would naturally want one$$and that the company$$s worldwide success and reputation would draw customers. But Google Glass did not consider important data that may have predicted the product$$s failure$$desire for privacy, the market for movie piracy, and properly functioning technology.

It$$s dangerous to assume correlation is the same as causation. Again, Google gives a great example with Google Flu Trends $($GFT$),$ implemented in $25$ countries. The company planned to predict flu trends based on searches for common flu symptoms like $$cough$$ and $$fever$.$ Google assumed that more searches using these terms during flu season would predict locations of outbreaks, but its predictions were off by as much as $140\%.$ The variables appeared to be predicting seasonal illness in general, not necessarily flu.

$4$b$.$ Google Flu Trends may have succeeded if the...

Google Flu Trends may have succeeded if the company had employed experts who had expertise needed in the analysis of data, based on