Lecture 1
- 1. PLAN FOR TODAY: Syllabus and Survey What is statistics (and why you should care) Different types of data In-class Exercise
- 2. A little about me: My survey answers (Q5 asks about your major): 1. United States (New York) 2. I’ve been studying/working with data for 20 years. 3. I lean towards Marvel with a few DC characters. 4. I’ve been using Excel for many years. 6. I went snowshoeing with my dog Bruno. 7. Every day, I am intrigued by how often data analysis is essential in all areas. Reasons why: 1) I am not usually here before class 2) I may not answer emails after 5:00 PM I grew up here (Syracuse) and went to SU and then I left for travel, work and education: Worked in Scotland, Slovakia, Lithuania, Chile Traveled all over… Went to graduate school in Oregon and Virginia Worked in West Virginia, Virginia, Mississippi, and Rhode Island Worked for federal gov’t and in the private sector; Still do consulting
- 3. PLAN FOR TODAY: Syllabus and Survey What is statistics (and why you should care) Different types and components of data In-class Exercise
- 4. WHAT IS STATISTICS, ANALYTICS, ETC? Simply put: Statistics allows us to answer questions about (almost) anything we want to know that we can collect data for. We collect a SAMPLE from our POPULATION to help answer Our questions Then we answer questions about our POPULATION based on our SAMPLE The answers we get from our sample are ESTIMATES of the Truth We can’t know the true value unless data are collected for whole Population
- 5. WHAT IS STATISTICS, ANALYTICS, ETC? Simply put: Statistics allows us to answer questions about (almost) anything we want to know that we can collect data for. We start with a POPULATION that we have a question about. We select a subset called a sample. Based on that sample we get an estimate, which is a STATISTIC That estimate, gives us an answer to our question.
- 6. WHAT IS STATISTICS? Statistics and analytics and data science allow us to answer questions about (almost) anything we want to know that we can collect data for. Examples: Politics: What percentage of people in the U.S. want the president to release his tax returns? Finance: Should I invest in Apple or Samsung based on their comparative past performance? Market research and Economics: How much do people spend on entertainment?
- 7. HOW MUCH DO WE SPEND ON ENTERTAINMENT, ANNUALLY? True answer (usually unknown) Can only be found by surveying the WHOLE POPULATION: • All civilian noninstitutional residents of the U.S (about 98% of total population) • In this case, this population is over 300 million people Can we collect date from all of them? • Not practical! Instead we collect data from a SAMPLE (subset) of the POPULATION: • 12000 households that are randomly selected (Consumer Expenditure Survey) • This SAMPLE is a subset that ideally represents the WHOLE POPULATION We use these sample data to estimate how much we spend on entertainment. • $652 (in 2015 *) * Data are from the U.S. Bureau of Labor Statistics (https://tinyurl.com/ybo5w6f6)
- 8. HOW MUCH DO WE SPEND ON ENTERTAINMENT, ANNUALLY? True answer (unknown) only found by surveying the WHOLE POPULATION: Can we collect data from all of them? • Nope! Not practical! Instead we get an estimate from a sample (subset) of the POPULATION: We collect data from our sample to estimate the true answer: • $652 (in 2015 *) • We have an estimate of the true answer! * Data are from the U.S. Bureau of Labor Statistics (https://tinyurl.com/ybo5w6f6) All civilian noninstitutional residents of the U.S More than 300 million people 12,000 households
- 9. PLAN FOR TODAY: Syllabus and Survey What is statistics (and why you should care) Different types and components of data In-class Exercise
- 10. Nation WTO Status Per Capita GDP Trade Deficit Fitch Rating Fitch Outlook Armenia Member 5,400 2,673,359 BB- Stable Australia Member 40,800 -33,304,157 AAA Stable Austria Member 41,700 12,796,558 AAA Stable Azerbaijan Observer 5,400 -16,747,320 BBB- Positive Bahrain Member 27,300 3,102,665 BBB Stable DATA SET Shown above is a subset of a data set with 60 rows The columns of the data set are VARIABLES Headers are variable names The rows of the data set are called OBSERVATIONS The individual values in the data set are DATA Fitch Rating: https://en.wikipedia.org/wiki/Fitch_Ratings Fitch Outlook: http://bit.ly/29vjjvf GDP: Gross Domestic Product - the total value of goods produced and services provided in a country during one year.
- 11. Nation WTO Status Per Capita GDP Trade Deficit Fitch Rating Fitch Outlook Armenia Member 5,400 2,673,359 BB- Stable Australia Member 40,800 -33,304,157 AAA Stable Austria Member 41,700 12,796,558 AAA Stable Azerbaijan Observer 5,400 -16,747,320 BBB- Positive Bahrain Member 27,300 3,102,665 BBB Stable DATA SET Shown above is a subset of a data set with 60 rows The columns of the data set are VARIABLES Headers are variable names The rows of the data set are called OBSERVATIONS The individual values in the data set are DATA Fitch Rating: https://en.wikipedia.org/wiki/Fitch_Ratings Fitch Outlook: http://bit.ly/29vjjvf GDP: Gross Domestic Product - the total value of goods produced and services provided in a country during one year.
- 12. TYPES OF DATA Categorical Data (Qualitative, Descriptive) Quantitative Data (Numerical, How much, How big, etc.) Ordinal (indicates rank or order) Nominal (no order, e.g. ID or company name) Discrete (whole numbers) Continuous (all real numbers e.g. decimals, fractions) Do not indicate magnitude or quantity Do indicate magnitude or quantity
- 13. TYPES OF DATA Nation WTO Status Per Capita GDP Trade Deficit Fitch Rating Fitch Outlook Armenia Member 5,400 2,673,359 BB- Stable Australia Member 40,800 -33,304,157 AAA Stable Austria Member 41,700 12,796,558 AAA Stable Azerbaijan Observer 5,400 -16,747,320 BBB- Positive Bahrain Member 27,300 3,102,665 BBB Stable Categorical Nominal Categorical Nominal Quantitative Continuous Quantitative Continuous Categorical Ordinal Categorical Ordinal
- 14. Last 4 Digits of SSN Credit Card Balance No. of Credit Cards FICO Rating Gender 3985 $501 1 VERY POOR Female 0954 $1,006 3 FAIR Male 4592 $201 7 VERY GOOD Female 3625 $964 4 GOOD Female TYPES OF DATA Categorical Nominal Quantitative Discrete Categorical Ordinal Quantitative Continuous Categorical Nominal (Some would Say ordinal, But I disagree)
- 15. Survey Respondent ID Number Total Years of Education Highest Degree Completed Gender 1 11 Left High School Male 2 16 Bachelor's Female 3 19 Graduate Female 4 12 High School Female 5 14 Junior College Male TYPES OF DATA Categorical Nominal Quantitative Discrete Categorical Ordinal Categorical Nominal (Some would Say ordinal, But I disagree)
- 16. Year Month Price of Domestic Crude Oil ($) 1973 June $3.56 1982 September $3,560 2009 August $71.20 2010 December $89.04 TYPES OF DATA Quantitative Discrete Categorical Ordinal Quantitative Continuous
- 17. School State Campus Setting Endowment ($ billions) Applicants Accepted (%) NCAA Division Amhearst MA Small Town 1.7 18 III Duke NC Midsize City 5.9 21 I-A Harvard MA Midsize City 34.6 9 I-AA Swarthmore PA Large Suburb 1.4 18 III TYPES OF DATA Categorical Ordinal OR Nominal Quantitative Continuous Categorical Nominal Categorical Nominal Quantitative Continuous (18% = 0.18) Categorical Ordinal
- 18. Summarizing Quantitative Data into Categories Sometimes we want to change a quantitative variable into a categorical one by summarizing the data. Why??? First 18 of 149 observations Hwy MPG Number of Cars 10-19 8 20-29 72 30-39 61 40-50 8 Total Number of Cars 149 Hwy MPG 30 29 31 28 27 19 24 22 30 29 28 24 24 29 29 38 37 36 This is a good first step in understanding the data How to do this (and many other things) are in the Excel tutorials In our next lecture, we’ll cover other ways to summarize quantitative variables
- 19. PLAN FOR TODAY: Syllabus and Survey What is statistics (and why you should care) Different types and components of data In-class Exercise
- 20. How many variables are in the data set FUELDATA2012 (in Blackboard ‘Lectures and Video’ – ‘Week 1 – Lecture 1’)? A. 7 B. 4 C.8 D.10 E. 3 The ID variable, car, is included when we count the total number of variables.
- 21. How many quantitative variables are in the data set, FUELDATA2012 (in Blackboard ‘Lectures and Video’ – ‘Week 1 – Lecture 1’)? A. 5 B. 4 C.8 D.3 ID variables, such as car, may appear quantitative, but they are nominal. They are used to identify each observation.
- 22. Which of the following variables in the data set FUELDATA2012 is both categorical and ordinal? A. Cylinders B. Size C.Drive D.Car Variables with text are always categorical. If the text values have an inherent order, then the variable is ordinal.
- 23. Which of the following variables in the data set FUELDATA2012 is both quantitative and discrete? A. Car B. Displacement C.Cylinders D.City MPG E. Hwy MPG If a quantitative variable refers to something countable, that can’t be subdivided, e.g., cylinders, people, etc., it is discrete.
- 24. True or false: city MPG and HWY MPG are both discrete variables. A. True B. False Variables like these may be rounded and appear discrete, but miles per gallon (MPG) can be a decimal value so this is a continuous variable.
- 25. Key points from today Population vs. Sample These concepts will be discussed throughout course Different terms for parts of a data set Columns are Variables Rows are Observations Individual values are Data 4 Main Types of variables Categorical Nominal (e.g. names, ID values) Categorical Ordinal (e.g Letter Grades, Quality Ratings) Quantitative Discrete (e.g., number of pets, years of education) Quantitative Continuous (e.g. credit card balance, height, weight) If you have a question (or two) about material from Lecture 1, please submit it by by the end of the day today. In Blackboard, go to Assignments > Lecture Questions > Lecture 1 Question(s)