Results chapter conducting, interpreting, and writing

Quantitative and Qualitative
Results Chapter 4:
Conducting, Interpreting, and Writing
by
Dr. James Lani

Results Chapter and
Threading a Needle
Threading a Needle

Conducting a Results Chapter

Once per sewed button

Do it once in your life

Not an end in itself (it’s to get a
button on to hold something
together)

It’s one part of one project in
one degree, in one part of your
life

You need the right ingredients:
needle, thread, and good eyes.

Software, knowledge, and time.

Right attitude: you don’t quit if
you can’t thread it 3 times.

Right attitude: you never give
up having come this far!

To schedule a consultation call 877-437-8622 or email Info@StatisticsSolutions.com

Today’s Webinar
• Quantitative Results
• Quantitative Results Strategy
• Data cleaning and Screening
• Descriptive statistics
• Conducting common analyses and reporting
findings
• Qualitative Results
• Qualitative Analysis Overview
• Grounded theory: Open, Axial, Selective coding
• Phenomenological: Lived experiences
• Q &A

Quantitative Results Strategy
•

•
•
•
•
•

Select the correct analysis (RQ and level
of measurement)
Clean your data
Describe variables
Conduct the analyses/assess assumptions
Present the findings
Summarize the findings


Data Cleaning and Preparation
• Assess data for outliers (±3.29);
• Multiple imputation for missing data;
• Create composite score (with reverse coding
if necessary);
• Conduct Cronbach’s alpha (α)


Descriptive Statistics
Means and Standard Deviations
Frequency and Percentages
Table 1
Frequencies and Percentages for Nominal Variables
Variables

n

%

72
78

48.0
52.0

36
43
71

24.0
28.7
47.3

Location
Urban
Rural

Ethnicity
White
Hispanic
Other


Conducting Frequently Used
Statistical Analyses
o
o
o
o
o
o
o
o
o

Chi-square: Goodness of fit and Test of independence
Pearson correlation
Independent samples t-test
One way ANOVA
Dependent samples t-test
Repeated-measures ANOVA
Linear regression
Multiple, Logistic, Ordinal, Multinominal regression
Mediation and Moderation


Chi-square: Goodness of Fit and
Test of Independence
• Chi-square analysis answers what research
question?
• Assumptions of analysis:
o Each cell has count of 1;
o 80% of cells have an expected value of 5.
• Conducting analysis;
• Presenting findings;
o Write-up in narrative;
o Tables and figures.

Chi-Square: Goodness of Fit
and Test of Independence
Republican

Democrat

Green

Independent

x

x

x

x

Republican

Democrat Green

Libertarian
x

Indep.

Libert.

Male

x

x

x

x

x

Female

x

x

x

x

x


Pearson Correlation
• Correlation analysis answers what research
question?
• Assumptions of analysis
Linearity, homoscedacity, normality
• Conducting analysis (e.g., IQ and Creativity)
• Presenting findings
Write-up in narrative
Tables and figures


Pearson Correlation Assumptions
Linearity

Non-linear


Homoscedasticit
y
Met

Hetroscedasticit
y


Normal

Non-Normal


Independent Samples t-test
Independent t-test analysis answers what research
question?
Males

Assumptions of analysis (IQ by
gender)
Homogeneity of variance
Normality
Conducting analysis

Presenting findings
Tables and figures

Females

Part 1=1

Part 4=2

Part 2=2

Part 5=3

Part 3=3

Part 6=4

X=2

X=3

Males

Females

Part 1=1.9

Part 4=2.9

Part 2=2.0

Part 5=3.0

Part 3=2.1

Part 6=3.1

X=2

X=3


One-Way ANOVA
ANOVA answers what research question?

Assumptions of analysis
Homogeneity of variance
Normality

Dem

Part 1=1.9

Conducting analysis

Repub

Part 4=2.9

Part 2=2.0 Part 5=3.0
Part 3=2.1

Presenting findings
Tables and figures

Indep

Part 6=3.1

X=2

X=3

X=


Dependent Samples t-test
Dependent sample t-test analysis answers what
research question?
Science
Pretest
Part 2=2

Part 2=3
Part 3=4

X=2

X=3

Science
Pretest

Science
Posttest

Part 1=1.9

Part 1=2.9

Part 2=2.0

Part 2=3.0
Part 3=3.1

X=2

Presenting findings
Tables and figures

Part 1=2

Part 3=2.1

Conducting analysis

Part 1=1
Part 3=3

Normality

Science
Posttest

X=3


Repeated-Measures ANOVA
Repeated-measures ANOVA answers what research
question?
Science
Pretest

Conducting analysis
Sphericity
(homogeneity of variance)

Science
Posttest

Science
Follow-up

Part 1=1.9

Part 1=2.9

Part 1=

Part 2=2.0

Part 2=3.0

Part 2=

Part 3=2.1

Part 3=3.1

Part 3=

X=2

X=3

X=

Presenting findings
Tables and figures


Linear Regression
Linear regression analysis answers what research question?
(e.g., Does IQ predict Creativity)
Normality, Multicollineality, Homoscedasticity
(residual vs. predicted values)
Conducting analysis

IV

DV

Presenting findings
Tables and figures

Regressions: Multiple, Logistic,
Ordinal, Multinominal

Its all about the level of measurement
of the DV


Mediation Analysis
Mediation analyses answers what kind of research
questions?
Assumptions of regression

Conducting analysis
Presenting findings
Tables and figures


Mediation Analysis Example
Education (M)

IQ (IV)

Creativity (DV)

3 Regression Equations

IVM; must be significant
IVDV; must be significant
M, IV  DV; IV is no longer significant

Moderation Analysis
Moderation analyses answers what kind of research
questions?
Assumptions of regression

Conducting analysis
Presenting findings
Tables and figures


Moderation Analysis
IQ (IV)
Age (Mod)
IQ x Age
Interaction

Creativity
Regression with 2 blocks
Step 1: IQ and Age entered
Step 2: Interaction term entered
Moderation is supported if interaction
is significant.

Note. To avoid multicollinearity, center IV/Mod (subtract mean), then
create the interaction term.


Qualitative Analysis Overview
• Sample size: no definitive requirement
• Grounded theory: Open, Axial, Selective coding (~20-30
participants)

• Phenomenological: Lived experiences
• (~5-25 participants)
• Inter-rater reliability: Kappa coefficient


Data Preparation
• Transcribe interviews
• Line number transcripts
• If thematizing without software, organize interviews
by interview question rather than interviewee;
• If you are using NVIVO software,
• Sources: Separate document by each interviewee;
• Node: Separated for each interview question,
themes, and subthemes.

Grounded Theory Example:
Effectiveness of Block Scheduling
Research Question 1: During the first year of block scheduling, how did
students perceive the effectiveness of the new class schedule?
Interview Question 1. “During the first year of block scheduling, how did you
feel about your class load?”

Open coding revealed two sets of theme related information based on the
level of difficulty and amount of information taught. (1) The majority of
participants cited the level of difficulty associated with block scheduling.
Level of difficulty was related to overall course load and class schedules. (2)
The amount of information taught in the classes was also mentioned by the
students.


Qualitative Analysis: Grounded
Table 1: Effectiveness Interview Question 1 Open Coding Responses
Theory

Part. Theme Phrases/descriptive words
Level of
difficulty

2

Initially it was very frustrating juggling so many classes, but eventually it got easier

3

In the beginning the classes were overwhelming, and there were problems, but
eventually I was able to manage

4

I think I had about the same amount of homework as before, so my classes are pretty
easy

5

It was easy and having a staggered schedule helped

6

The tests were a lot harder because we were able to cover more information in a shorter
amount of time, and we were tested on all of it

8

In the beginning it was hard, but for the second half of the year I had figured out how to
manipulate my schedule so I had an even work load most of the week

10

It was fine, pretty easy transition

Axial coding for feelings about the class load revealed the following
Theory
thematic relationships (in order from strongest to weakest associations):
Four of the ten (40.0%) specified more information was taught/covered
(e.g., covered chapters a lot faster, not used to learning this much, more
information in a shorter amount of time).
Three of the ten (30.0%) specified the year started out difficult but got
easier (e.g., frustrating but got easier, overwhelming but eventually I could
manage, and hard but eventually I figured it out).
Three of the ten (30%) specified the class load was pretty easy (e.g., classes
are easy, easy transition, and having a staggered schedule made it easier).


Theory
Selective coding of interview question 1, feelings about the class load,

generated a composite response of block scheduling was easy once an
understanding was gained (e.g., frustrating but got easier, overwhelming
but eventually I could manage, fine, pretty easy transition, and hard but
eventually I figured it out), followed by the theme of more information was
taught (e.g., covered chapters a lot faster, not used to learning this much,
more information in a shorter amount of time).
The researcher selected one participant’s statement that
reflected the axial coding composite with the most congruence.
Participant 3 stated, “In the beginning the classes were overwhelming, and
there were problems, but eventually I was able to manage. Having the block
schedule allows me to have a study period, so even though the tests cover
more information now I can be prepared because I have more time to
study.”

Qualitative Analysis:
Phenomenological
• Read entire set of transcripts to get a feel
for what is being said;

• Read across each interview question for all
participants, pulling themes out of
passages and then quoting (excerpts)


Qualitative Analysis: Kappa
Rater 1
Coefficient
Rater 2

For
Present
Theme 1

Present

Absent

Subtotal

A

B

A+B

C

D

C+D

A+C

B+D

A+B+C+D

Absent

Subtotal

Observed agreement = (A + D)
Expected Agreement = (((A + B) * (A + C)) + ((C + D) * (B + D))) / (A + B + C + D)
Kappa = ((observed agreement) – (expected agreement)) / ((A + B + C + D) –
(expected agreement))
Note: A, B, C, and D are the frequencies in which a theme is identified in same
excerpt between rater 1 and rater 2.


Qualitative Analysis References
Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing
among five traditions. Thousand Oaks, CA: Sage Publications.
Glaser, B. G. & Strauss, A. L. (1967). The discovery of grounded theory:
Strategies for qualitative research. Piscataway, New Jersey: Transaction.
Morse, J. M. (1994). Designing funded qualitative research. In Denizin, N. K. &
Lincoln, Y. S., Handbook of qualitative research (2nd Ed). Thousand
Oaks, CA: Sage.
Moustakas, C. (1994). Phenomenological research methods. Thousand
Oaks, CA: Sage.
Patton, M. Q. (1990). Qualitative evaluation and research methods (2nd Ed.).
Thousand Oaks, CA: Sage.


All Yours:
Questions & Answers

Personalized 1-1
Quantitative and
Qualitative Consulting
Email:
Info@StatisticsSolutions.com
877-437-8622

We’re interested…
Webinar Feedback and
What You’d Like
Please provide feedback on the webinar and tell us what else
you’d like to learn about in upcoming webinars.
Send comments to: James@StatisticsSolutions.com

Thank You
for your Participation
and Attention!
Join us for our next webinar on
Wednesday, January 22nd at
8:30pm ET

Results chapter conducting, interpreting, and writing

More Related Content

What's hot (12)

Viewers also liked (14)

Similar to Results chapter conducting, interpreting, and writing (20)

More from Statistics Solutions (20)

Recently uploaded (20)

Results chapter conducting, interpreting, and writing