INTRODUCTION
CHAPTER 1
OVERVIEW
• Descriptive and Inferential Statistics
• Variables and Constants
• Levels of Measurement
• Discrete and Continuous Variables
• The NELS data set
• Interfacing with RStudio
Statistics: Descriptive and Inferential
• Descriptive: describe the data that have
been (or will be) collected.
• Inferential: generalize for a population
from a small randomly selected sample
VARIABLES AND CONSTANTS
Variables – Vary from person to person or
object to object
Constants – Remain constant from person to
person or object to object
EXAMPLE
A study is conducted to determine if there are gender
differences in fine motor skills among five year-olds from
middle class families.
• What are the variables in the study?Gender, Fine motor skills
• What are the constants in the study?Age, Socio-economic status
Note: Whether a trait is a constant or a variable is determined
by the nature of the particular study.
LEVELS OF MEASUREMENT
• Nominal
• Ordinal
• Scale
• Ratio
• Interval
NOMINAL
Objects observed to be similar on some characteristic
(e.g., college student) are assigned to the same class or
category, while objects observed to be dissimilar on
that characteristic are assigned to different classes or
categories. (Example: Car types: Toyota, Honda, Volvo,
BMW, Audi, etc.)
ORDINAL
The ordinal level of measurement is not only based on
observing objects as similar or dissimilar, but also on
ordering those observations in terms of an underlying
characteristic. (Example: Height in size place ranking)
the heights of each pair of
adjacent people will not
necessarily also differ by the same
value.
SCALE
An ordinal level of measurement can be developed into a
higher level of measurement if it is possible to assess how
near to each other the persons or objects are in the
underlying characteristic being observed, that is, if numbers
can be assigned in such a way that equal numerical
differences along the scale correspond to equal increments in
the underlying characteristic.
If this is done so that the score 0 corresponds to having none
of the property being measured, the level of measurement is
ratio. (Example: Distance from home to school) Otherwise, it
is interval. (Example: Temperature Scale)
OTHER EXAMPLES
Find the levels of measurement of the following variables
measured on college students:
• Residence: 1 = College dormitory, 2 = Off campus apartment, 3 = Parents’
home, 4 = Other Nominal
• Whether or not the student’s parents live in the same state as the student’s
college Nominal or Ordinal
• Class rank Ordinal
• Year graduated high school Scale: Interval
Scale: Ratio
• Number of earned credits to date
• Height Scale: Ratio
• Family Income Scale: Ratio
• Number of siblings Scale: Ratio
DISCRETE AND CONTINUOUS
VARIABLES
• A variable is discrete if the values it takes on are integers
or can be thought of in some unit of measurement in which
they are integers.
• A variable is continuous if in any unit of measurement,
whenever it can take on the values a and b, it can also
theoretically take on all the values between a and b.
EXAMPLES
Classify each of the following variables as discrete or continuous:
• Residence: 1 = College dormitory, 2 = Off campus apartment, 3 = Parents’
home, 4 = Other Discrete
• Whether or not the student’s parents live in the same state as the student’s
college Discrete
• Class rank Discrete
• Year graduated high school Discrete
• Number of earned credits to date Discrete
• Height Continuous
• Family Income Continuous
• Number of siblings Discrete
This is an RStudio working session which displays four windows in a single screen. The console
window is on the bottom left, and the .R file window, containing a set of commands from Ch 2, is
on the top left. In addition, the top right window provides three tabs (Environment, History,
Connections) for displaying objects in working memory available within those tabs. For example,
within the Environment tab, one may obtain a listing of the datasets you have loaded into working
memory for that session. Finally, the bottom right window contains five tabs (Files, Plots,
Packages, Help, Viewer) that when clicked displays the information relevant to that tab. In this
figure, because the Help tab is shown to have been clicked and because ?hist had been
typed into the console during this R session, a description of the histogram command
appears in this window.
THE NELS DATA SET
• The NELS data set was collected by the U.S. Department of
Education’s National Center of Education Statistics (NCES).
• It is a nationally representative longitudinal data set to measure
achievement outcomes in four core subject areas (English,
history, mathematics, and science), in addition to personal,
familial, social, institutional, and cultural factors that might
relate to these outcomes.
WHAT IS THE LEVEL OF MEASUREMENT
FOR THE FOLLOWING VARIABLES FROM
THE NELS DATA SET?
• Twelfth grade math achievement (achmat12). Scale: Interval
• Whether or not the student took advanced math in eighth
grade (advmat08). Nominal or Ordinal
• Urbanicity (urban). Ordinal
• region.
Nominal
• The number of hours spent on hw in school in twelfth grade
(hwkin12) – Look carefully at the coding! Ordinal
• Similar variables are hwkout12, late12, cuts12, absent12,
extracurr12.
Scale: Interval
• Socio-economic status (ses).
• Rating of teacher interest (tchrint). Scale: Interval