Lecture 3a Notes

Qualtrics

• Used to gather data, functioning like a “net” to collect responses.

• Data can be exported to SPSS (preferred for this course) or Excel for analysis.

• SPSS is used in this class to enhance your CV, as it is a less common skill than Excel, but valuable to employers.

Exporting Survey Text

• You can export the survey questions to a Word document using Tools > Export survey to Word.

• Include a copy of your survey in the appendix of your project report.

Data and Analysis options

• Select Data and Analysis to access data collected from the survey.

Question Coding and Recoding

• Qualtrics typically codes answers with numerical levels (1, 2, 3, etc.).

• It's possible to change the values using the recode values option in Qualtrics.

• Recoding makes it easier to analyze data, as numerical values can represent different categories or levels.

• When you export Qualtrics data to SPSS or Excel, Qualtrics replaces the question response with a number, so 'less than one' might become 'one' (the first option).

• Example: Recode cities to numerical values (1 = Melbourne, 2 = Sydney, 3 = Brisbane) for concise data entry and analysis.

Data Types

• Nominal data: Labels without inherent order (e.g., city of residence).

• Ordinal data: Ordered labels (e.g., HD, D, C, P, Z), where the order matters.

• Recoding is crucial when the default alphanumeric order doesn't make sense.

Example of Recoding

• If grades (HD, D, C, P, Z) are initially ordered alphabetically (C, D, F, HD, P), recode them to reflect the correct order of achievement (e.g., 1 = HD, 2 = D, 3 = C, 4 = P, 5 = Z).

Matrix Tables and Likert Scales

• With matrix tables, each row is technically a separate question (e.g., rating multiple brands on the same scale).

• Qualtrics exports matrix table data with appended numbers (e.g., Q111 for Revlon, Q112 for Maybelline).

Scaling

• Likert scales typically range from "strongly disagree" to "strongly agree."

• Consider reversing the scale (positive to negative) for better user experience, but be aware some researchers change the scale to ensure respondents are paying attention.

Testing Surveys

• Test surveys with friends or colleagues to check clarity and data validity before deployment.

• You can generate “faux responses” to simulate data and test statistical outputs in SPSS before deploying to a real audience.

How Data Is Stored

• Data is stored as labels or numerical indicators (1, 2, 3…).

Exporting Data to Excel

• Options include CSV, TXT, Excel, XML, and SPSS formats.

• For Excel, you can export labels or values, but not both simultaneously.

Limitations of Excel

• Analyzing labels in Excel is difficult, while analyzing numbers may cause you to lose sight of the meanings of labels.

Data Types and Measures of Central Tendency

• Nominal data: Only mode (most common value) can be calculated.

• Ordinal data: Mode and median can be calculated.

• Numeric data: Mode, median, and mean can be calculated.

• Measures of central tendency: Mode, median, and mean.

• Each data type determines what can be measured.

Measures of Central Tendency

• Mode, median, and mean are all measures of central tendency.

• Measures inform data such as the average age of survey respondents.

Introduction to SPSS

• SPSS allows you to have both labels and numerical values.

• When saving files, use syntax to keep a record of all manipulations, which helps in case of mistakes.

• Use a naming convention such as underscore raw for raw data files.

Syntax

• .SPS is the extension for an SPSS syntax file.

• Syntax skills are valuable and enhance your CV.

Download Options in Qualtrics

• Leave the width at medium for text fields (limits text field to 2,000 characters).

• You can untick to prevent Qualtrics from compressing the file into a .ZIP.

• .SAV is the notation for an SPSS data file

Using SPSS

• After downloading, double-click the file to open it in SPSS.

• The data view and variable view allows you to navigate data

SPSS Interface

• The syntax file will automatically label each question for you.

Variable and Value Labels

• Variable labels are displayed in variable view.

• Value labels show what each numerical code represents (e.g., 1 = “less than one”).

• Value labels and variables display together in SPSS.

Using SPSS for Data Analysis

• Nominal values can be automatically displayed when clicking on value labels.

• SPSS allows switching between viewing nominal values and value labels.
  Variable names can be customized in SPSS for easier analysis. It's useful to concatenate the question number with the shortened variables (e.g. Q2_TypeOfHair)