Unit I Test

(TPS) Chapters 1 & 2

(WS2) topics 1, 2, 7, 8, 9, 10,12
A. Data
1. Identify the individuals (observational units) and variable in a set of data.
2. Identify each variable as categorical (binary or not) or quantitative (discrete or
continuous). Identify the units(ft, in, etc) in which each quantitative variable is
measured.
B. Displaying Distributions
1. Make/interpret a bar graph of the distribution of a categorical variable.
2. Represent categorical data using frequency or relative frequency(%) tables.
3. Make/interpret a graph of the distribution of a quantitative variable:
i. dotplot for a fairly small set of observations without a large range.
ii. stemplot also for a fairly small set of observations. Round leaves or split
stems as needed to make an effective stemplot.
iii. histogram, usually for large sets of data.
iv. boxplot (or modified boxplot). Typically used for comparing distributions.
v. Cumulative graph. Shows the number or proportion of a data set less
than or equal to the given number. Be able to find percentiles, median,
and quartiles from a cumulative relative frequency graph.
C. Inspecting/describing Distributions (Quantitative variables only) (SOCS+Context)
1. Look for the overall pattern and for major deviations from the pattern.
2. Assess from a dotplot, stemplot, boxplot, or histogram whether the shape of a
distribution is roughly symmetric, distinctly skewed right or left, or neither.
Assess whether the distribution has one or more major peaks (unimodal,
bimodal) or is approximately uniform (no peaks – flat). All of these graphs will
allow you to see shape, but a boxplot will not show peaks or whether the data is
mound shaped.
3. Describe the overall pattern by giving numerical measures of
i. Center – (mean, median)
ii. Spread – (range, IQR, standard deviation)
iii. Decide which measure of center and spread are more appropriate: the
mean and standard deviation (typically for symmetric distributions) or
the five number summary: min, Q1, Med, Q3, max (typically for skewed
distributions)
4. Recognize outliers.
i. Outliers can be identified visually on most graphs by a fairly significant
gap between the highest or lowest observations and the rest of the data.
ii. Calculate outliers using the 1.5XIQR rule. (Draw modified boxplots)
iii. Calculate outliers using the empirical rule and +/- 2 or 3 standard
deviations.
 Unit I Test
(TPS) Chapters 1 & 2
(WS2) topics 1, 2, 7, 8, 9, 10,12
D. Measuring Center(Mean/Median)
1. Find the mean (𝑥̅ ) of a set of observations. Be sure you can understand tally
tables.
2. Find the median of a set of observations. Make sure you know the difference
when the data set is even or odd.
3. Understand that the median is more resistant (less affected by extreme
observations) than the mean.
4. Recognize that skewness in a distribution moves the mean away from the
median toward the tail.
i. Symmetric: mean = median
ii. Skewed right: mean > median
iii. Skewed left: mean < median
E. Measuring Spread (Range, IQR, Standard Deviation)
1. Find the five-number summary (min, Q1, Med, Q3, max) and draw a boxplot.
2. IQR is the only resistant measure of spread. Use range and IQR with median.
3. Using a calculator, find the standard deviation (s) for a set of observations. (1-Var
Stats) If using mean, use st. dev. for spread.
4. Know the basic properties of standard deviation:
i. 𝑠 ≥ 0 always (it can’t be negative)
ii. 𝑠 = 0 only when all observation are identical
iii. s increases as the spread increases
iv. s has the same units as the original measurements
v. s is not resistant (increased by outliers or skewness)
vi. standard deviation is the ‘typical’ distance the data values all fall from the
mean. (an estimate of the avg. distance that all of the points fall from the
mean)**Know the definition and be able to write it in context.**
F. Changing Units of Measurement (linear transformations)
1. Determine the effect of a linear transformation on measures of center and
spread.
i. If you add or subtract a constant to all of the data:
1. Measures of center are increased or decreased by the same
amount.
2. Measures of spread are not affected.
ii. If you multiply or divide a constant to all of the data:
1. Measures of center are increased or decreased by that same
factor.
2. Measures of spread are also increased or decreased by that same
factor.
 Unit I Test
(TPS) Chapters 1 & 2
(WS2) topics 1, 2, 7, 8, 9, 10,12
G. Comparing Distributions
1. Use side-by-side bar graphs to compare distributions of categorical data.
2. Make back-to-back stemplots and side-by-side boxplots to compare distributions
of quantitative variables.
3. Write narrative comparison of the shape, center, spread, and outliers (SOCS) for
two or more quantitative distributions. Be sure to use comparative words:
Greater, similar, less, etc. Don’t just describe the two distributions separately.
H. Percentiles
1. interpreted as the value that has that % of the data less than or equal to it. Be
able to find and interpret the percentile of an individual value in context.
2. Estimate percentiles and individual values using a cumulative relative frequency
graph.
I. Normal Curves: (Mound shaped/symmetric)
1. Know the empirical rule and how to estimate the amount of data above or below
1, 2, or 3 standard deviations from the mean.
i. Approx.. 68% of the data in a normal curve is within one st. dev. of the
mean.
ii. Approx.. 95% of the data in a normal curve is within two st. dev. of the
mean.
iii. Approx.. 99.7% of the data in a normal curve is within three st. dev. of the
mean.
2. Know how to calculate a z-score (standardized score).
𝑥−𝜇
i. 𝑧 = where 𝜇 is the mean and 𝜎 is the standard deviation of the
𝜎
population (parameters).
ii. Interpret a z-score as the number of st. deviations a value is above (+) or
below (-) the mean.**Be able to write in context.**
3. Find the proportion of values in a specified interval in a Normal distribution using
Table A or technology.
4. Find the value that corresponds to a given percentile in a Normal distribution
using Table A or technology.
5. Determine whether a distribution of data is approximately Normal from
graphical and numerical evidence.
i. Mound shaped/symm/follows the empirical rule
ii. Looking at a normal probability plot. The normal probability plot is/isn’t
fairly linear, so the data set (in context) is/isn’t approx. normal.