Professional Documents
Culture Documents
Prior to Analysis
What would you like to discover?
Perceived competence
Preferences, satisfaction
Group differences
Demographics
Prior to Analysis
Your goals drive the make-up of the survey
analysis plan.
Prior to Analysis
Survey design & layout
Survey Design
Good questions reduce error
something consistently?
supposed to be measuring?
Construct Validity
Does your measure correlate with a theorized
concept of interest?
Pilot
Piloting the survey can inform:
Question clarity
Question format
Variance in responses
Survey Analysis
Using data from
Paper Surveys
SurveyMonkey
SelectSurvey.Net
Survey Analysis
Paper surveys
Survey Analysis
Paper surveys
Survey Analysis
SurveyMonkey
Available under the analyze results tab
Survey Analysis
SelectSurvey.NET
Available under Analyze Results Overview
Frequencies
Download all responses for further analysis
Select Export Data from Analyze page
Export Format CSV (excel)
Data Format SPSS Format Condensed
Data Cleaning
Process of detecting, diagnosing, and editing
faulty data
Basic Issues:
lack or excess of data
outliers, including inconsistencies
unexpected analysis results and other types of
inferences and abstractions
Data Cleaning
Inspect the data
Frequency distributions
Summary statistics
Graphical exploration of distributions
Scatter plots, box plots, histograms
Data Cleansing
Out of range
Delete values and determine how to recode if possible
Missing Values
Refusals (question sensitivity)
Dont know responses (cant remember)
Not applicable
Data processing errors
Questionnaire programming errors
Design factors
Attrition
Missing Data
Missing completely at random (MCAR)
Cases with complete data are indistinguishable from
cases with incomplete data.
Missing at random (MAR)
Nonignorable
The pattern of data missingness is non-random and it
is related to the missing variable.
Missing Data
Listwise or casewise data deletion: If a record has missing
Descriptive Statistics
Frequency distribution
Descriptive Statistics
Cross-tabs
Excel Pivot tables
Data Analysis
Measurement scale determines how the data
should be analyzed:
Data Analysis
Three instructors are evaluating preferences
Data Analysis
Nominal & ordinal variables are discrete
Can be qualitative or quantitative
Interval & ratio variables are continuous
Grades
Age
Data Analysis
Charts
Pie charts & bar charts
used for categorical
data
Histograms used for
continuous data
Line graphs typically
show trends over time
Data Analysis
Other descriptive statistics
Mean
Median
Mode
nominal data
Data Analysis
Other descriptive statistics
Interquartile range
Standard deviation
Correlations
Are the variables related?
Correlation
Which correlation coefficient to use?
Pearsons r
Mean Differences
Are there meaningful differences between
groups?
class sections
instructors
on-line vs. off-line courses
major vs. non-major
Mean Differences
Which test to run?
Interval and ratio data
Presenting Results
Describe the purpose of the survey
List the factors that motivated you to conduct
this research in the first place.
Include the survey!
On assessment reports
When the survey is new/still being fine tuned
Presenting Results
Present the breakdown of results
Tables and graphs should complement text
Conclusions
Recommendations
Describe an action plan based on concise
concluding statements
Presenting Results
Share results in formal venues
Familiarize yourself with key findings so that
Moving Forward
Continuously improve the survey
Delete, add, change questions
Evaluate method of administration
Compare results across semesters to look for
improvements
Compare with other assessment data for a
broader picture