Detailed Tutorial for using the features of this class Just click your way through the course. Here
are all the Web lectures and learning activities for this class listed
in the order you need to do them. Each lecture has a set of activities
designed to help you learn that topic. Just click on each
activity in the order they are listed and you'll move along right
through the course.
By clicking and working on the activities in sequence you will learn the fundamental concepts of statistics
in interesting and innovative ways.
Navigating to Other Course Features. Interactive, Notes,
Practice, Grades, Communicate>, Resources and Take
Exams to go to those class features.
There is no homework for this topic. But it is a very important
topic that gives meaning to how statistics fits with scientific
thinking. These ideas will be parts of several required web homeworks
as well as several recommended practice homeworks. These ideas will
be tested on both midterm and final exams.
Use of the Normal Tool is necessary to do the required web homework
as well as the recommended practice homework. Instructions for using
the Normal Tools are integrated into the Normal Distribution lecture
(click on Get Knowledge).
Using the Sample from Normal Tool is really part of the Normal
Distribution topic, but it requires some thoughtful integration
of ideas. The Sample from Normal Tool allows you to take random
samples from a normal probability distribution. Every sample is
different, so every student gets different data and consequently
different answers. So there is no "correct" answer. Do
not look for an answer key; there is none. What is important is
to understand the concept.
It is very important to give your self EXPERIENCE with sampling, since
the process of sampling is crucial to understanding many statistical concepts
that we will cover. It's not particularly hard to understand, all you
need to do is have some experience with sampling. Come back and play with
the Sample from Normal Tool later in the class to reinforce these ideas
when we are using them.
The Normal Tool is also used for a few problems in the Binomial
Homework by way of review.
Note (If you have trouble manipulating windows).
To do Binomial Homework, we recommend the following procedure. (1)
Click on "Do Homework" on the list above to open the Web
Homework. Select Binomial Distribution from the web homework menu.
(2) Click on "Use Binomial Tool," above. Two
pages open--the tool itself (pale yellow) and a web page (brown,
with the StatCenter symbol). Use the Binomial Tool to do all the
Web Homework problems asking about the binomial distribution. (3)
When you get to homework problems that ask you about the Normal
Distribution, click "Back" on the (brown) binomial web
page. Ducks in a Row will appear in its place. (4) Click
on "Normal Tool" in the above list. Again two pages open--the
normal tool itself and a (brown) web page. Use the Normal Tool to
complete the web homework. Submit the homework. Click "Back"
to see Ducks in Row again. (On small screens you may have to minimize
windows during the above procedure to see the homework and tools
clearly.)
We will use the Binomial distribution as a foundation for important
ideas later; so it is important to start learning bout it. Do the
homework; it will guide toward making important conclusions, But,
since the Binomial Tool is interactive, you can play around with
it, making your own discoveries and getting your own sense of the
Binomial distribution. When playing change N and p. Change between
probability inside and probability outside. Just follow your curiosity
long enough to be familiar with how the Binomial works.
Up to this point we have been developing a lot of theory. Now you'll
start learning some statistics. We'll come back and use the theory
after a while, when you've had time to digest it.
Central Tendency (the Mean, the Median, and the Mode) are pretty
straightforward. Things get a little more complicate as learn about
how to measure and describe how "spread out" a group of
numbers are.
The Double Sample Tool allows you to set up two Normal Distributions,
one red and one green. You can make them as close together or as
far apart by changing each of their mu's. You can make them as compact
or spread out as you want by changing their sigma's. Then you can
sample from both of them simultaneously. You get descriptive statistics
for each sample. If you pay attention to the data in the two samples
you gain invaluable experience about how differences in data sets
indicate differences in populations.
The Detect Difference & Double Sample lecture is one of the
most important lectures in this course as far as gaining
a deep understanding of the meaning of the statistics. Please read
it carefully and make good notes. Understanding this material
will make later parts of the course much easier.
When you click on "Use double sample tool" a menu comes
up. Just click on "Double Sample" to use the tool.
Below tool's menu are two problems (scroll down). Problem 1 gives
you a structured experience with the Double Sample Tool. But it
is important to play around and interact with the tool, making your
own conclusions.
Come back to this tool when we are learning about a t-test for
independent means later; this tool will give you insight into that
statistic.
Detect Difference simulates the basic puzzle faced by scientists
when they do a two group study. Suppose a scientist has two groups
of volunteer participants. S/he gives one group a placebo and the
other group a newly developed chemical that might lower blood pressure.
No one knows if it lowers blood pressure or not; that's why the
scientist is doing research--to find out. At the end of the study
s/he has two samples of numbers, one sample from each group.
S/he looks at the numbers. If the new chemical is ineffective (as
most are) then what s/he is looking at is two samples drawn from
the same population. If the new chemical is effective at lowering
blood pressure then the two samples are drawn from two different
populations. (That is, if the chemical works, then the group which
received the chemical is a sample from a population of people with
low blood pressure--and the placebo group is a sample from a population
with high blood pressure.)
S/he looks at the two sets of numbers. S/he has to decide whether
these two samples lead to the conclusion that there are two populations
(the chemical is effective) or that there is only one population
(the chemical is ineffective).
This game puts you into that scientific puzzle. You have the data.
Do you conclude there is one population or two populations? That's
the essence of a deep and common scientific puzzle.
If you play around you will notice that there buttons with little
labels like "t=", "M", "SD", and "SEM".
These are statistics that you've not learned about. But if you press
the buttons you get the statistics even if you don't know what they
mean. You can press the buttons or not. If you do you may notice
they are useful to you are not. Probably the mean (M) will be useful.
Either way (whether the stats make sense or not) it's ok. You're
just gaining experience with a scientific thinking puzzle.
Of course, how well you do is recorded and counts toward your course
grade.
In the second part of the Regression
Lecture you will learn about prediction errors, prediction error variance,
and the idea of least squared error. These are important and deep
concepts which underlie the philosophy of statistics. Consequently
they are a bit harder to learn. Expect that it is normal to have to
study this material over a few times to get a good sense of it.
Interact
and Integrate: Virtual Lab, Assignment #1 Virtual Lab, Assignment #1 Chapter #1:
| Current Score: N/A
Virtual Lab, Assignment #1 Chapter #2:
| Current Score: N/A
Virtual Lab, Assignment #1 Chapter #3:
| Current Score: N/A
Get
Knowledge: Virtual Lab Tutorial
Make Notes (Not Available) Do Virtual Lab Assignment
#1 (Three Chapters Required, Each Lab can be repeated multiple times)
Do Homework (Homework is built into Virtual Lab: No other homework)
Virtual Lab is a highly innovative program that lets you simulate
the whole scientific problem solving process from theories and hypotheses
to designing research studies to collecting and analyzing data to
making conclusions and writing up reports.
This is Assignment #1 for Virtual Labs. Please choose only Volume
4: "Explorations" as
your book to read for Assignment #1. Also, you must do
three (3) different research projects (Chapters) to complete Assignment
#1. Read "Get Knowledge: Virtual Lab Tutorial" for more details.
"Difference to Inference" is a game that brings together
many threads of learning that you've encountered up to this point.
More importantly, it provides you will experiences that will greatly
help in understanding theoretical concepts that we will develop
later. "Difference to Inference" builds on your experience
with "Double Sample" and "Detect Difference."
It gives you an opportunity to use Sample Data drawn from Normal
Distributions to make Scientific Inferences about which of several
theories is the most scientifically viable.
Difference to Inference Grade: At this time your assignment
is to play the EASY and the MEDIUM levels of the Difference to Inference
Game. (Later, after we learn about t-tests, you will be required
to play the HARD level.) You must earn 2000 or more Grant Bucks
on each game level (Easy or Medium) to get to get a grade of 100%.
If you get less than 2000 grant bucks, your grade will be the percentage
of 2000 that you earn. If you get more than 2000 grant bucks, you
still only get 100%. But there is a list of the top 10 researchers
in the class available to look at; so there is prestige (such as
it is) attached to earning more grant bucks.
Sampling Distributions are a pivotal concept in the true understanding
of statistical theory. The lectures, homeworks, and tools developed
for this course are holistically integrated to provide you with
the direct kind of experiences that enable you to understand these
elusive ideas.
Estimating
population parameters | Current Score: N/A
We have established the distinction between populations and samples.
In the mathematical-statistical model, that data we collect in a
research is a sample from a probability distribution (which is called
a population). We are now going to learn how to make guesses (estimates)
about the parameters of a populations from sample data.
You have now completed all the material tested on the midterm
exam.
You should continue to learn new material at the beginning of the
next week. This new material will not be covered on the Midterm
but it is crucial that you keep up with the pace of the course.
Online Students: Sign up for Midterm Exams.
Midterm
Exam Form 1 | Current Score: N/A
Online: Register for Section 90 Midterm Exam Form 1
You are required to make arrangements to find and coordinate with an
approved proctor (a university or college testing center) to administrate
your exam(s). To take an exam, you will need to submit an Exam Request
Form through the Distance Education office. Keep in mind that this coordination
between Distance Education and your approved proctor could take time,
so proper planning is essential. In addition, there are costs associated
with proctors and amounts vary depending on the institution. If you have
any questions regarding exams, including approved proctors, please contact
the Distance Education office at (800) 467-8839 or email at distance@aoce.utah.edu.
NOTE: You must pass the final exam in order to pass the course. If you
fail this exam you will fail the course. This is a requirement of all
U of U Distance Education courses.
Hypothesis Testing is a set of concepts in the formal mathematical-statistical
model. Statistical Conclusion Validity is a logical issue in the
philosophy of science and research methods. Statistical Conclusion
Validity deals with the question, "Could my results (data)
have occurred by chance alone?" Scientists use the formal mathematical
model to evaluate the role of chance in determining the outcome
of their research.
In the Sampling Distributions, Estimating Parameters, and Hypothesis
Testing Lectures we laid out the theoretical foundations of inferential
statistics. The t-independent Lecture develops a powerful inferential
statistic with wide applicability.
The t-correlated Lecture continues the development of common uses
of the the t statistic. It also introduces an important measurement
distinction--are the measures independent of each other or are they
correlated with each other? Consider one simple example. Suppose
you are evaluating the effectiveness of a Psychotherapy. Case #1:
You could have two (independent) groups of participants. One group
would receive Psychotherapy and then be measured for mental health.
The other group would receive a Placebo Control and then be measured
for mental health. In other words, you have two groups, and each
group is measured only once. In contrast is Case #2. You could have
only one group of participants and measure them twice for mental
health, once before Psychotherapy and once after Psychotherapy.
Case #1 has two groups, each measured once. Case #2 has one group,
measured twice. The measurements in Case #1 are consdered independent
and the appropriate t-test is t-independent. The measurements in
Case #2 are correlated with each other and the appropriate t-test
is t-correlated. There is more to say on this matter, and you will
find a more extensive discussion in the t-correlated Lecture.
The t for r allows you evaluate whether a correlation coefficient
(r) is significantly different than 0.
Midterm
Form 2
| Current Score: N/A
Midterm Form 2 is optionally available at the end of this week.
Take Form 2 only if you want to improve your grade. If you take
both forms you will automatically receive the higher of the two
grades.
Online students: (optional) Register for Section 90 Midterm Exam
Form 2
You are required to make arrangements to find and coordinate with an
approved proctor (a university or college testing center) to administrate
your exam(s). To take an exam, you will need to submit an Exam Request
Form through the Distance Education office. Keep in mind that this coordination
between Distance Education and your approved proctor could take time,
so proper planning is essential. In addition, there are costs associated
with proctors and amounts vary depending on the institution. If you have
any questions regarding exams, including approved proctors, please contact
the Distance Education office at (800) 467-8839 or email at distance@aoce.utah.edu.
NOTE: You must pass the final exam in order to pass the course. If you
fail this exam you will fail the course. This is a requirement of all
U of U Distance Education courses.
The t for b statistic allows you to evaluate whether or not a regression
slope (b) is significantly different than 0. The t for b and the
t for r tests are formally the same test. If calculated on the same
data, they will give you the same value.
Interact
and Integrate Use Virtual Lab (Three Chapters Required, Each Lab can be repeated multiple times)
Virtual Lab Assignment #2: In the first Virtual Lab assignment
you only used descriptive statistics. Now you have learned enough to do
a more sophisticated Virtual Lab assignment. This assignment will allow
you to practice using all the t-tests in an open-ended context. You will
have to decide which of the t-tests to use and demonstrate your ability
to use them properly. You will also be required to make statistical conclusions
about the "significance" (statistical conclusion validity) of
your results. The most difficult discrimination is when to use t-independent
versus t-correlated. In fact, one of the motivating reasons for developing
Virtual Lab was because students demanded lots of open-ended practice
in telling when to use t-correlated and when to use t-independent so they
would be ready for the test and so they would really understand the research
issue involved. Please read only Volume 5 for
this assignment. You must do three (3) research puzzles from Volume
5. (It is good practice for the final exam to do more puzzles;
but only three count for a grade.)
Getting to StatTool. Once you have collected data in Virtual
lab, just close the data clipboard and you will see a button that
says "stat tool." Press that button. A page will come
up (sometimes slowly) with an overview of instructions for using
StatTool on it. It also has a large "Analyze Data" button.
Look over (or print) the instructions and then press "Analyze
Data." You will see a pale yellow "Raw Data" window
with your Virtual Lab data in it. Next to the raw data window is
a blank white "Statistical Results" window; later, when
you do a data analysis the results of you statistical analysis will
appear in this white window. Both the white and the pale yellow
windows can be resized in case all the data or all the statistical
results don't fit the default window sizes.
Take notes. Take notes on the statistical results that appear
in the white Statistical Results window. These will not import back
into Virtual Lab. So you will need to jot down the analyses so that
you can use them when you make a conclusion for your Virtual Lab
research project.
Chi-squared
Goodness of Fit Get
Knowledge Make
notes
Chi-Square Goodness of Fit homework (Web Homework not available)
Practice (Recommended)
Here we introduce another common and useful inferential statistic.
Chi-Square in widely used to evaluate the PCH of Chance when we
have categorical frequency data. Therefore, in this lecture, we
introduce an important distinction--the difference between measurement
data and frequency data.
The Chi-Square Test of Association allows you to determine if there
is a significant relationship between two categories such as Politcal
Party Affiliation and Enviromental Attitudes.
One
way ANOVA independent measures | Current Score: N/A
People speak of "one-way" or "one-factor" ANOVA's.
ANOVA is short or "Analysis of Variance." "One-factor"
or "one-way" simply mean "one independent variable."
ANOVA is a very flexible system that can analyze many IV's at once,
We begin with the simplest case: The one-way ANOVA for independent
groups.
The one factor Analysis of Variance (ANOVA) for independent groups
extends the t-independent statistical procedure from two groups
to any number of groups.
The two way ANOVA for independent groups allows us to analyze research
projects that manipulate two IV's at the same time. An important new concept
introduced in these lectures is the idea of an interaction between two
IV's. Continue to work on the material two-way ANOVA for independent groups.
Especially focus on the concept of interaction.
Review
and prepare for the exam. Complete all coursework Online: Register for Final Exam Form 1
You are required to make arrangements to find and coordinate with an
approved proctor (a university or college testing center) to administrate
your exam(s). To take an exam, you will need to submit an Exam Request
Form through the Distance Education office. Keep in mind that this coordination
between Distance Education and your approved proctor could take time,
so proper planning is essential. In addition, there are costs associated
with proctors and amounts vary depending on the institution. If you have
any questions regarding exams, including approved proctors, please contact
the Distance Education office at (800) 467-8839 or email at distance@aoce.utah.edu.
NOTE: You must pass the final exam in order to pass the course. If you
fail this exam you will fail the course. This is a requirement of all
U of U Distance Education courses.
You are required to make arrangements to find and coordinate with an
approved proctor (a university or college testing center) to administrate
your exam(s). To take an exam, you will need to submit an Exam Request
Form through the Distance Education office. Keep in mind that this coordination
between Distance Education and your approved proctor could take time,
so proper planning is essential. In addition, there are costs associated
with proctors and amounts vary depending on the institution. If you have
any questions regarding exams, including approved proctors, please contact
the Distance Education office at (800) 467-8839 or email at distance@aoce.utah.edu.
NOTE: You must pass the final exam in order to pass the course. If you
fail this exam you will fail the course. This is a requirement of all
U of U Distance Education courses.