STAT 241 Principles of Statistics: Data Collection and Summarization

Topic 1: Data Collection and Summarization

Introduction to Statistics

Statistics is the science of collecting, summarizing, analyzing, and interpreting data. It provides methods for making sense of data and drawing conclusions about populations based on samples.

• Data: Collections of observations, such as heights, ages, or scores.

• Statistics: The discipline concerned with data collection, summarization, analysis, and interpretation.

• Applications: Used in fields such as medicine, economics, sports, and social sciences.

Populations, Samples, Census, Parameters, and Statistics

Understanding the distinction between populations and samples is fundamental in statistics.

• Population: The entire group of individuals or items under study (e.g., all USI students).

• Sample: A subset of the population selected for analysis (e.g., students in STAT 241).

• Census: Data collected from every member of the population.

• Parameter: A fixed, unknown numerical value describing a population characteristic (e.g., population mean $\mu$, variance $\sigma^2$).

• Statistic: A numerical value describing a sample characteristic, used to estimate parameters (e.g., sample mean $\overline{X}$, sample variance $s^2$).

Types of Data

Data can be classified as categorical or quantitative, and further as discrete or continuous.

• Categorical Data: Data representing categories or labels (e.g., car model, marital status).

• Quantitative Data: Data representing numerical values or counts (e.g., number of students, time spent).

• Discrete Variable: Takes only countable values (e.g., number of baskets made).

• Continuous Variable: Can take any value within a range (e.g., height, weight).

Levels of Measurement

Measurement levels determine the type of statistical analysis that can be performed.

• Nominal: Names, labels, or categories without order (e.g., zip code).

• Ordinal: Data can be ordered, but differences are not meaningful (e.g., course grades).

• Interval: Ordered data with meaningful differences, but no true zero (e.g., temperature).

• Ratio: Interval data with a true zero; differences and ratios are meaningful (e.g., weight, length).

Types of Studies

Studies can be observational or experimental, each with distinct purposes and methodologies.

• Observational Study: Passive data collection without influencing subjects.

• Experimental Study: Imposing treatments to study effects; includes experimental and control groups.

• Cross-sectional Study: Data collected at one point in time.

• Retrospective Study: Data collected from past records.

• Prospective Study: Data collected in the future from cohorts.

Sampling Methods

Sampling is crucial for obtaining representative data from populations.

• Random Sampling: Every element has an equal chance of selection.

• Simple Random Sampling: Every possible sample of size n has an equal chance.

• Systematic Sampling: Select every k-th element after a random start.

• Convenience Sampling: Use readily available elements.

• Stratified Sampling: Divide population into strata and sample from each.

• Cluster Sampling: Divide population into clusters, randomly select clusters, and sample all members.

• Multistage Sampling: Combine multiple sampling methods in stages.

Errors in Data Collection

Errors can arise from sampling or from data collection and analysis.

• Sampling Error: Difference between sample result and true population result due to chance.

• Nonsampling Error: Errors from incorrect data collection, recording, or analysis.

Summarizing Data: Tables and Graphs

Data can be summarized and visualized using various tables and graphs.

• Frequency Table: Shows categories (classes) and their frequencies.

• Histogram: Bar graph for quantitative data; bars represent frequency.

• Scatter Plot: Graph of paired (x, y) data to show relationships.

• Time Series Plot: Data plotted over time.

• Dot Plot: Each data value is a dot along a scale.

• Stem-and-leaf Plot: Data split into stem and leaf for visualization.

• Bar Graph: Bars represent categories of qualitative data.

• Pareto Chart: Bar graph with bars in descending order of frequency.

• Pie Chart: Circle divided into slices proportional to frequency.

Descriptive Statistics: Measures of Center

Measures of center describe the typical value in a data set.

• Sample Mean ($\overline{X}$): Arithmetic average of sample data. Formula: $\overline{X} = \frac{\sum_{i=1}^n x_i}{n}$

• Population Mean ($\mu$): Arithmetic average of population data. Formula: $\mu = \frac{\sum_{i=1}^N x_i}{N}$

• Weighted Mean: Mean where data values have different weights. Formula: $\overline{X} = \frac{\sum_{i=1}^n w_i x_i}{\sum_{i=1}^n w_i}$

• Median: Middle value when data are ordered; if even number, average the two middle values.

• Mode: Most frequently occurring value in the data set.

• Quartiles: Divide data into four equal parts (Q1, Q2, Q3).

• Percentiles: Divide data into 100 equal parts.

Descriptive Statistics: Measures of Variation

Measures of variation describe the spread or dispersion of data.

• Range: Difference between largest and smallest data values. Formula: $\text{Range} = \text{Max} - \text{Min}$

• Interquartile Range (IQR): Difference between third and first quartiles. Formula: $\text{IQR} = Q_3 - Q_1$

• Sample Variance ($s^2$): Average squared deviation from the mean. Formula: $s^2 = \frac{\sum_{i=1}^n (x_i - \overline{X})^2}{n-1}$

• Sample Standard Deviation ($s$): Square root of variance. Formula: $s = \sqrt{s^2}$

• Population Variance ($\sigma^2$): Average squared deviation from the population mean. Formula: $\sigma^2 = \frac{\sum_{i=1}^N (x_i - \mu)^2}{N}$

Chebyshev's Theorem

Chebyshev's theorem provides a minimum proportion of data within k standard deviations of the mean for any data set.

• At least $1 - \frac{1}{k^2}$ of data lies within k standard deviations of the mean, for $k > 1$.

• For $k = 2$, at least 75% of data within 2 standard deviations.

• For $k = 3$, at least 89% of data within 3 standard deviations.

The Empirical Rule

For bell-shaped (normal) distributions, the empirical rule gives approximate percentages of data within 1, 2, and 3 standard deviations of the mean.

• About 68% within 1 standard deviation.

• About 95% within 2 standard deviations.

• About 99.7% within 3 standard deviations.

Coefficient of Variation (CV)

The coefficient of variation expresses the standard deviation as a percentage of the mean, useful for comparing variability between data sets.

• Sample CV: $\text{CV} = \frac{s}{\overline{X}} \times 100\%$

• Population CV: $\text{CV} = \frac{\sigma}{\mu} \times 100\%$

Z Score

The z score indicates how many standard deviations a data value is from the mean.

• Sample z: $z = \frac{x - \overline{X}}{s}$

• Population z: $z = \frac{x - \mu}{\sigma}$

• Example: For $x = 7$, $\overline{X} = 5$, $s = 3.3$, $z = \frac{7-5}{3.3} = 0.61$

Five-Number Summary and Boxplot

The five-number summary provides a concise description of a data set, and the boxplot visualizes it.

• Five-number summary: Minimum, Q1, Median (Q2), Q3, Maximum.

• Boxplot construction:

  1. Find the five-number summary.

  2. Draw a scale including min and max values.

  3. Draw a box from Q1 to Q3, with a line at the median.

  4. Draw whiskers from the box to min and max values.

• Outliers: Data values more than $1.5 \times \text{IQR}$ above Q3 or below Q1.

HTML Table: Frequency Table Components

The frequency table is a key tool for summarizing quantitative data. Below is a summary of its main components:

Component	Description	Example
Lower Class Limit (LCL)	Smallest value in a class	10 in class 10-19
Upper Class Limit (UCL)	Largest value in a class	19 in class 10-19
Class Boundary	Midpoint between LCL and UCL of consecutive classes	19.5 between 10-19 and 20-29
Class Midpoint	Middle value of a class	14.5 for class 10-19
Class Width	Distance between consecutive LCLs	10 for classes 10-19, 20-29

Shapes of Distributions

Histograms can reveal the shape of data distributions.

• Skewed Right (Positive Skew): Long tail to the right; many small values.

• Skewed Left (Negative Skew): Long tail to the left; many large values.

• Symmetric: Bell-shaped; data evenly distributed around the mean.

Other Graphs for Data Visualization

Various graphs help visualize different types of data.

• Scatter Plot: Shows relationship between two quantitative variables.

• Time Series Plot: Shows data over time.

• Dot Plot: Plots each data value as a dot.

• Stem-and-leaf Plot: Splits data into stems and leaves for visualization.

• Bar Graph: Visualizes categorical data frequencies.

• Pareto Chart: Bar graph with descending frequencies.

• Pie Chart: Shows proportions of categories as slices of a circle.

Summary Table: Measures of Center and Spread

Measure	Formula (LaTeX)	Purpose
Mean	$\overline{X} = \frac{\sum_{i=1}^n x_i}{n}$	Central tendency
Median	Middle value (ordered data)	Central tendency, robust to outliers
Mode	Most frequent value	Central tendency for categorical/quantitative data
Range	$\text{Range} = \text{Max} - \text{Min}$	Spread of data
Variance	$s^2 = \frac{\sum_{i=1}^n (x_i - \overline{X})^2}{n-1}$	Average squared deviation
Standard Deviation	$s = \sqrt{s^2}$	Typical deviation from mean
Interquartile Range	$\text{IQR} = Q_3 - Q_1$	Spread of middle 50% of data
Coefficient of Variation	$\text{CV} = \frac{s}{\overline{X}} \times 100\%$	Relative variability
Z Score	$z = \frac{x - \overline{X}}{s}$	Standardized value

Example: Five-Number Summary and Boxplot

• Sample: 1, 3, 5, 2, 7, 4, 5, 5, 13, 5

• Sorted: 1, 2, 3, 4, 5, 5, 5, 5, 7, 13

• Five-number summary: Min = 1, Q1 = 3, Median (Q2) = 5, Q3 = 5, Max = 13

• Boxplot: Box from Q1 to Q3, line at median, whiskers to min and max.

• Outlier check: Any value above Q3 + 1.5 × IQR or below Q1 − 1.5 × IQR is an outlier.

Additional info: These notes cover foundational concepts in statistics, including data types, sampling, descriptive statistics, and graphical methods, which are essential for further study in probability, distributions, and inferential statistics.