Data Science – Fall 2016

MIS 0855 – Prof. Min-Seok Pang

Exam #1 grades and Unofficial LETTER Grades posted on Gradebook

I have posted Exam #1 grades and unofficial LETTER grades on MIS Community Gradebook at http://community.mis.temple.edu/gradebook.

The unofficial letter grades are strictly unofficial and only to inform what you would get if this is the end of the semester. Be advised that your final grades depend on your performance for the rest of the semester.

If you would like to take a look at your Exam #1, please drop by my office hours next Monday (Oct 10) or Thursday (Oct 13).

In-Class Exercise – Week 05 – Finding Cool Data Visualization

In-Class Exercise – Week 05 – Finding Cood Data Visualization.pdf

Examples of not-so-interesting hypotheses

Hypothesis 1. “CSI is a better television show than Law and Order.”

Hypothesis 2. “MLB All Star game was more entertaining back in the 1990s than it is in the 2010s.”

Hypotheses 1 and 2 are not testable. It needs to be more specific in terms of measurement. How can we measure about “better television show”? How to measure the degree of “entertaining”?

Hypothesis 3. “The cost of living is cheaper in Philadelphia than New York.”

Hypotheses 1-3 are not interesting because they compare only a set of two facts (e.g. cost of living in Philadelphia and New York, quality/popularity of CSI and Law and Order). Instead of comparing the cost of living in two cities, a better hypothesis could be “the cost of living is higher at a city whose annual average temperature is between 70F and 80F than at other cities.” By doing so, we compare cost of living and temperatures in multiple cities.

Hypothesis 4. “A city with higher inches of rainfall has more rainy days in a year.”

Hypothesis 4 is too obvious. You’ll get better grades from hypotheses whose rationale could be counter-intuitive and surprising.