Mulitvariate Analysis

1
Mulitvariate Analysis
Brian Cunningham, Ina Gibson, Ben Lee Handler,
Dalena Hunter and Linda Rivas
2
The greatest evil is physical pain. Multivariate
analysis -Saint Augustine
Most real world phenomena are multi-factorial in
nature, meaning there is more than one factor
that impacts on, or causes changes in the
dependent variable. In order to predict the
dependent variable as accurately as possible, it
is usually necessary to include multiple
independent variables in the model. Multiple
linear regression allows us to test how well we
can predict a dependent variable on the basis of
multiple independent variables.
From Statistically Significant Consulting for
Doctoral Students and Researchers
3
Multivariate analysis is.

• Real life
• Focused on one or more DEPENDENT VARIABLES (Y)
  and
• The relationship of INDEPENDENT VARIABLES
  (X1,X2) and the DEPENDENT VARIABLE Y

4
Three Main Functions

• Control
• Elaboration
• Prediction

5
How Does this Magic Happen?

• ltInsert REALLY UGLY Formula Heregt
• (Seriously, you will learn about the techniques
  and methods used later in the presentation so pay
  attention)

6
An Example

• Grad School Admissions
• Y Success in school
• X GRE scores
• Writing ability (admissions essay)
• Undergraduate performance
• Age, gender
• Red Hair
• Sleep patterns
• Pet growing up?

7
Can You Spot the Imposters?

• Spurious Relationships
• Red Hair
• Sleep patterns
• Pet Ownership

8
Sorting out the rest

• Identifying the Relative strengths of
  Independent Variables
• Identifying Additional, Intervening Variables
• What other variables account for the effect?

9
This stuff is important

• Hey, they have their own journal!

(And I think it is written in English!)
10
When do you say, enough!?

• In theory, there are no limits on the number of
  variables you can test for.
• In practice, you will experience AUTOCORRELATION,
  adding new variables will not yield enough
  information to justify the time and effort needed
  to gather all the extra data.

11
A Final Word of Advice

• DO NOT TRY THIS AT HOME
• (At least not without a computer
• and a good software package)

12
Benefits of Multivariate Analysis

• The three basic functions of multivariate
  analysis- control, interpretation, and prediction
  allow the researcher to fully examine the
  relationship between the dependent variable and
  one or more independent variables.

13
Benefits of Multivariate Analysis

• Control- Multivariate Analysis allows the
  researcher to control association between
  variables by using cross tabulation, partial
  correlation and multiple regression.

14

Benefits of Multivariate Analysis

• Elaboration- Multivariate analysis allows the
  researcher to elaborate on the relationship
  between two variables in order to determine the
  relationship between the variables or to
  determine the conditions necessary to create a
  specific relationship between variables.
  Introduces other variables to determine the links
  between the independent and dependent variables
  or to specify the conditions under which the
  association takes place

15
Benefits of Multivariate Analysis

• Elaboration
• Intervening variable- link between the
  independent variable and the dependent variable
• Interaction- conditions or contingencies
  necessary for the relationship to occur
• Interest and concern as a condition-
  differences in interest affect subject attitude
  and behavior patterns
• Time and place as a condition- relationship
  between two variables can vary according to the
  time and place at which it is studied.
• Background Characteristics as a condition-
  associations may differ for persons or groups
  who do not share the same characteristics

16
Benefits of Multivariate Analysis

• Prediction (Path Analysis?)- uses bivariate and
  multiple linear regression to test the causal
  relations among the variables specified in the
  model.
• 3 steps researcher draws a path diagram based
  on a theory, researcher calculates path
  coefficients, researcher determines indirect
  effects.

17
Benefits of Multivariate Analysis

• Multiple Relationships (I think this is a con as
  it is confusing to non math people)
• Multivariate analysis allows researchers to study
  the combined effect of all independent variables
  on the dependent variable.

18
IT Takes a Long Time
19
The Math is HARD
20
Its Expensive
21
Its Complicated
22
ITS Difficult to Display
23
In the End, We Still Dont Know a Whole Lot
24
Nahyun Kwon, Anthony J. Onwuegbuzie, and Linda
Alexander.Critical thinking disposition and
library anxiety Affective domains on the space
of information seeking and use in academic
libraries. College Research Libraries 68, no3
268-78, May 2007.
25
Critical Thinking
26
Library Anxiety
27
What is the relationship between ones critical
thinking disposition and ones feeling of library
anxiety?
28
California Critical Thinking Disposition
Inventory (CCDTI)You know you needed another
acronym in your life.
29
Strong critical thinkers are truth seeking,
open-minded, analytical, systematic,
self-confident (well at least in matters of
critical thinking), inquisitive, mature (please
dont giggle).
30
Library Anxiety not referring to the feelings
associated with being an MLIS student, but rather
the recurring fear and feeling of being lost
among students who use an academic library for
their research. (Kwon 121)
31
Library Anxiety was measured using the Library
Anxiety Scale (LAS -- Another one!!!)
32
Symptoms include shortness of breath (a patron
complains that they need to take a couple of
deep breaths,) fears of appearing foolish and
general uneasiness.
33
Findings revealed a negative multivariate
relationship, meaningstudents with weak
critical thinking dispositions tend to have high
anxiety levels, and conversely, a strong
disposition meant far lower anxiety levels.
34
The findings led to the authors to make a final
recommendation The library staff needs to
enable patrons to have positive critical thinking
dispositions so that they will reduce their
amount of library anxiety.
35
Bibliography

• Nachmias, Chava and David Nachmias. 2000.
  Control, Elaboration, and Multivariate Analysis.
  In Research Methods in the Social Sciences. 6th
  ed. New York Worth Publishers.
• Kranzler, Gerald and Janet Moursund.1999.
  Statistics for the Terrified. 2nd ed. New Jersey
  Prentice Hall.
• Nahyun Kwon, Anthony J. Onwuegbuzie, and Linda
  Alexander.Critical thinking disposition and
  library anxiety Affective domains on the space
  of information seeking and use in academic
  libraries. College Research Libraries 68, no3
  268-78, May 2007.