Econ 3790: Business and Economics Statistics Instructor: Yogesh Uppal Email: yuppal@ysu.edu

Lecture 1 — Schedule Goals of the course Data and statistics Tabular methods for summarizing data Graphical methods for summarizing data

Why use Statistics? To make sense of large amounts of data: What are the demographics of Youngstown in 2000? Have U.S. wages increased since 1975? To test hypotheses: Is demand curve downward sloping? Are GDP and Saving Rate positively correlated? To make predictions: What might happen to savings behavior after a large tax cut? These are specific to economics, but there are of course example from other disciplines: Medicine – drug testing These are increasing in their complexity. At the end of Econ 10A, you may be able to answer the first two on your own. These are univariate studies – one variable at a time. The others require looking at many variables at one time, and often require advanced methods.

Data: Basic Definitions Data: a set of measurements Dataset: all data collected for one study Element, or unit: an entity on which data are collected Variable: a property or attribute of each unit Observation: the values of all variables for one unit

Data: Basic Definitions Variables Observation Element Names Stock Annual Earn/ Exchange Sales($M) Share($) Company Dataram EnergySouth Keystone LandCare Psychemedics AMEX 73.10 0.86 OTC 74.00 1.67 NYSE 365.70 0.86 NYSE 111.40 0.33 AMEX 17.60 0.13 Data Set

Data: Scales of Measurement Four scales of measurement: Nominal, ordinal, interval, and ratio scales Scale determines which methods of summarization and analysis are appropriate for any given variable These are different ways that variables can be measured. Binary: taking exactly one of two possible values

Data: Scales of Measurement Characteristic Nominal, like a label or name for a characteristic e.g., color: red, green, blue race: black, Hispanic, white, Asian binary: (male, female), (yes, no), (0, 1) Ordinal, still a characteristic, but having a natural order e.g., how was service?: poor, average, good These are different ways that variables can be measured. Binary: taking exactly one of two possible values

Data: Scales of Measurement Numeric Interval scale Numeric data showing the properties of ordinal data e.g., SAT scores, Fahrenheit temperature Ratio scale Ordered, numeric data with real zero e.g., income, distance, price, quantity http://www.math.sfu.ca/~cschwarz/Stat-301/Handouts/node5.html

Data: Other Classifications Qualitative, or categorical: measures a quality Quantitative: numeric values that indicate how much or how many Cross-sectional: data collected at one point in time Time series: data collected over several time periods Panel or longitudinal: combination of cross-sectional and time series

Data: Summary of Definitions Qualitative Quantitative Numerical Nonnumerical Numerical Nominal Ordinal Nominal Ordinal Interval Ratio

Statistical Inference: Definitions Population: the set of all elements of interest in a study Sample: a subset of the population Statistical Inference: the process of using data obtained from a sample to make estimates and test hypotheses about the characteristics of a population

Statistical Inference: Process 1. Population consists of all tune-ups. Average cost of parts is unknown. 2. A sample of 50 engine tune-ups is examined. 3. The sample data provide a sample average parts cost of $79 per tune-up. Draw diagram with sample inside of the population, then with an arrow showing statistical inference. 4. The sample average is used to estimate the population average.

Descriptive Statistics: Definition Descriptive statistics are the tabular, graphical, and numerical methods used to summarize data

Descriptive Statistics: Common Methods Some common methods: Tabular Frequency table (for one variable) Crosstabulation, or crosstab (for more than one variable) Graphical Bar graph (for categorical variables) Histogram (for interval- or ratio-scaled variables) Scatterplot (for two variables) Numerical Mean (arithmetic average) This is certainly not a complete list, but you should feel comfortable starting this class knowing that you’ve actually seen or even used most of these.

Summarizing Qualitative Data Frequency distribution Relative frequency distribution Bar graph Pie chart Objective is to provide insights about the data that cannot be quickly obtained by looking at the original data

Distribution Tables Frequency distribution is a tabular summary of the data showing the frequency (or number) of items in each of several non-overlapping classes Relative frequency distribution looks the same, but contains proportion of items in each class Define classes: can be categories, or intervals of continuous data. Open stata, and use it with the census dataset to make frequency distns and rf distns .

Example 1: What’s your major? Frequency Finance 6 Marketing 10 Accounting 18 Advertising

Summarizing Quantitative Data Frequency Distribution Relative Frequency Distribution Dot Plot Histogram Cumulative Distributions

Example 1: Go Penguins Month Opponents Rushing TDs Sep SLIPPERY ROCK 4 NORTHEASTERN at Liberty 1 at Pittsburgh Oct ILLINOIS STATE at Indiana State WESTERN ILLINOIS 2 MISSOURI STATE at Northern Iowa Nov at Southern Illinois WESTERN KENTUCKY 3

Example 1: Go Penguins Rushing TDs Frequency Relative Frequency 3 0.27 3 0.27 1 2 0.18 0.09 4 0.37   Total=11 Total=1.00

Example 2: Rental Market in Youngstown Suppose you were moving to Youngstown, and you wanted to get an idea of what the rental market for an apartment (having more than 1 room) is like I have the following sample of rental prices

Example: Rental Market in Youngstown Sample of 28 rental listings from craigslist:

Frequency Distribution To deal with large datasets Divide data in different classes Select a width for the classes Need rules for constructing classes because the classes are not obvious with quantitative data

Frequency Distribution (Cont’d) Guidelines for Selecting Number of Classes Use between 5 and 20 classes Datasets with a larger number of elements usually require a larger number of classes Smaller datasets usually require fewer classes

Frequency Distribution Guidelines for Selecting Width of Classes Use classes of equal width Approximate Class Width = If classes are not equal width, you should be sceptical that someone is trying to deceive you.

Frequency Distribution For our rental data, if we choose six classes: Class Width = (750-330)/6 = 70

Relative Frequency To calculate relative frequency, just divide the class frequency by the total Frequency Remind them not to worry about percent frequency because it’s just relative times 100

Relative Frequency Insights gained from Relative Frequency Distribution: 32% of rents are between $539 and $609 Only 7% of rents are above $680

Histogram of Youngstown Rental Prices So explain how frequency tables are exactly the data contained in a histogram. In fact, when stata calculates the data for a histogram, it is in fact creating a frequency or relative frequency table.

Describing a Histogram Symmetric Left tail is the mirror image of the right tail Example: heights and weights of people .05 .10 .15 .20 .25 .30 .35 Relative Frequency

Describing a Histogram Moderately Left or Negatively Skewed A longer tail to the left Example: exam scores .05 .10 .15 .20 .25 .30 .35 Relative Frequency

Describing a Histogram Moderately Right or Positively Skewed A longer tail to the right Example: hourly wages .05 .10 .15 .20 .25 .30 .35 Relative Frequency

Describing a Histogram Highly Right or Positively Skewed A very long tail to the right Example: executive salaries .05 .10 .15 .20 .25 .30 .35 Relative Frequency

Cumulative Distributions Cumulative frequency distribution: shows the number of items with values less than or equal to a particular value (or the upper limit of each class when we divide the data in classes) Cumulative relative frequency distribution: shows the proportion of items with values less than or equal to a particular value (or the upper limit of each class when we divide the data in classes) Usually only used with quantitative data!

Example 1: Go Penguins (Cont’d) Rushing TDs Frequency Relative Frequency Cumulative Fre. Cumulative Relative Fre. 3 0.27 1 2 0.18 5 0.45 0.09 6 0.54 7 0.63 4 0.37 11 Total  

Cumulative Distributions Youngstown Rental Prices

Crosstabulations and Scatter Diagrams So far, we have focused on methods that are used to summarize data for one variables at a time Often, we are really interested in the relationship between two variables Crosstabs and scatter diagrams are two methods for summarizing data for two (or more) variables simultaneously

Crosstabs A crosstab is a tabular summary of data for two variables Crosstabs can be used with any combination of qualitative and quantitative variables The left and top margins define the classes for the two variables

Example: Data on MLB Teams Data from the 2002 Major League Baseball season Two variables: Number of wins Average stadium attendance

Crosstab Frequency distribution for the wins variable for the attendance variable

Crosstabs: Row or Column Percentages Converting the entries in the table into row percentages or column percentages can provide additional insight about the relationship between the two variables

Crosstab: Row Percentages

Crosstab: Column Percentages

Crosstab: Simpson’s Paradox Data in two or more crosstabulations are often aggregated to produce a summary crosstab We must be careful in drawing conclusions about the relationship between the two variables in the aggregated crosstab Simpsons’ Paradox: In some cases, the conclusions based upon an aggregated crosstab can be completely reversed if we look at the unaggregated data

Crosstab: Simpsons Paradox Frequency distribution for the wins variable Frequency distribution for the attendance variable

Scatter Diagram and Trendline A scatter diagram, or scatter plot, is a graphical presentation of the relationship between two quantitative variables One variable is shown on the horizontal axis and the other is shown on the vertical axis The general pattern of the plotted lines suggest the overall relationship between the variables A trendline is an approximation of the relationship

Scatter Diagram A Positive Relationship: y x

Scatter Diagram A Negative Relationship y x

Scatter Diagram No Apparent Relationship y x

Example: MLB Team Wins and Attendance Slightly positive relationship, or no relationship indicated

Tabular and Graphical Descriptive Statistics Data Qualitative Data Quantitative Data Tabular Methods Graphical Methods Tabular Methods Graphical Methods Freq. Distn. Rel. Freq. Distn. Crosstab Bar Graph Pie Chart Freq. Distn. Rel. Freq. Distn. Cumulative Freq. Distn. Cumulative Rel. Freq. Distn. Crosstab Histogram Scatter Diagram Ogive is just a plot of the cumulative distribution