Blog

Match the six steps of the analytics-based decision-making process. 

AdServe is an in-app video advertising company. The company offers a platform that embeds video ads in mobile apps to encourage users to download and install additional apps. When a user launches an app, his or her device sends a request to AdServe for an ad. For instance, suppose Malcolm is playing the game Sonic Dash. AdServe’s platform determines the best ad to serve to Malcolm while he plays Sonic Dash. Suppose AdServe decides to serve Malcolm an ad for the game Hay Day. Assuming Malcolm is still playing Sonic Dash when the ad is served, he sees the video for Hay Day. If Malcolm is interested in learning more about Hay Day, he clicks on the ad and is redirected to the app store. He might then decide to install Hay Day. On a typical day, using its current ad-serving algorithm (Algorithm A), AdServe experiences a 5% click-through rate. Analyzing consumer data, AdServe has realized that small improvements in click-through rates could have a large effect on the firm’s revenue. Realizing this, the company has developed a new ad-serving algorithm (Algorithm B) that could potentially improve the click-through rate. The company now wants to test whether the click-through rate of the new algorithm is significantly different from that of the existing algorithm. As is typical in such experiments, the two conditions, A (AdServe’s existing algorithm) and B (the new algorithm) would be evaluated in parallel on randomly assigned users.     a. (5 points) Develop a null and an alternative hypothesis to examine whether the performance of the new algorithm (Algorithm B) is different than that of the existing algorithm (Algorithm A) in affecting the click-through rate.       b. (5 points) You used an independent sample t-test to test the hypotheses. Below is the test result. Do you think AdServe should adopt the new algorithm? Why or why not?  Average CTR of algorithm A: 0.0492 Average CTR of algorithm B: 0.0506 t = – 1.030 p-value = 0.3030  

In a regression output, the t-value of a regression coefficient tells whether the corresponding independent variable affects the dependent (outcome) variable at a given significance level (say, at the 5% significance level).

What was the purpose of the pig drawing exercise?

Given the information below, what would you recommend the firm do to make capacity management easier?

There are two t-tests (test A and test B) that examine the consumption amounts of two different consumer groups (e.g., male consumers and female consumers). Due to random sampling, the sample of consumers selected for test A is generally different from the sample of consumers selected for test B. Suppose in test A, the difference between the means of the two groups selected for the test (i.e., test A) is 10, while in test B, the difference between the means of the two groups selected for the test (i.e., test B) is just 1. If test B concluded that the two group means were statistically different at the 5% significance level, then test A will conclude that its two group means are also statistically different at the 5% significance level.

The below diagram is an example of what?

Singapore Post followed the Service Recovery Framework properly.

An overdose of protamine may result in a hypercoagulable state.

 Consider the equation used to determine inhaled anesthetic uptake by blood from the lungs. Which factor has no significant impact on anesthetic uptake into the blood?