Stat 11 Week 1
What are Data and Statistics?
Prof. Suzy Thornton
1. What is data?
If not data, what else?
Structure of data
Observational units - any individual (human or otherwise) upon which we are collecting information in a systematic way.
In general, we usually say we have a sample of size \(n\) observational units, where \(n\) can be an integer greater than or equal to one.
Variables - a measurable or categorizable quality that tends to varies from one observational unit to another.
Type of variables
- Numeric/Quantitative variables are measurable and typically have some units of measurement associated with them
- Categorical/Qualitative variables which you may not be able to
measure but the outcomes are distinct enough to label according
to different qualities.
- Context matters determine what type of variable is being used
Dwight: We need to talk!
Michael: Not now.
Dwight: Which is higher? Assistant Regional Manager or Regional Director in Charge of Sales?
Michael: I told you the titles are irrelevant. They just relate to pay scale.
Dwight: Okay. So who gets paid more? Me or Andy?
Michael: It is not a matter of more or less. Your pay is just different. Okay?1
E.g. Age, time of day, or satisfaction level could all be treated as numeric or categorical variables depending on the discretion of the data analyst.
2. What is statistics?
Machine learning, computer science, data science, demography, modeling, mathematics, natural science, liberal art?
Models
What does the concept of a model mean to you?
Probability is a mathematical construct which defines a function called a random variable. In statistics, we use these mathematical functions (random variables) to model the behavior of a variable in a data set.
Models (including random variables) exist in the mathematical world. Data exists in the real world. Statistics uses tools from the mathematical world to investigate data that is carefully observed and draw conclusions about huge (but often unobservable) populations that exist in the real world.
Exploring vs confirming
Statistical analyses can be used to explore data or to draw conclusions about the strength of observable evidence in support of (or against) a proposed theory.
For example…
An economist wants to determine if there is evidence for a neo-classical business cycle within the global paper industry.
A CEO wants to get a sense of which of his store branches are consistently performing well or not.
A branch manager needs to determine if the proportion of faulty printers being sold by the company is larger than \(5\%\).
Employees have experienced a lot of changes in management and think that this is related to changes in their productivity.
What is a question you would like to answer with statistics? Is it exploratory or confirmatory?
3. About this class
The Textbook and Pearson MyLab are required. Attendance is required.
Everything you need for this class can be found on Moodle!
Analyzing data
You will be using software (either Excel or RStudio) to analyze data. There is an Intro to R (and RStudio) two-part workshop being held by the Stat Clinicians. Bring your own laptop! (Let me know if you need to borrow one from the department.)
Syllabus
By yourself: Take a few minutes to read over the syllabus.
What stands out to you?
What do you think is going to be the most important thing you can do to succeed in this class?
What do you think will be your biggest obstacle?
With someone else:
Share your answers with a neighbor. Strategize how to handle what you each expect will be your biggest obstacles.
Prof Suzy’s Recommendations
Read your textbook!
Practice asking and answering questions in class and while you read your textbook or do your homework.
Attend Office Hours as often as possible, this time is for you so use it!
Communicate with your professor. Keep me in the loop!
Allow yourself room to grow. We learn and live at our own paces. You’re not perfect and you don’t have to be.