Lecture 3D Flashcards (2)

Descriptive Summary Statistics

• How to Obtain Descriptive Statistics in SPSS:

  • Bring up your dataset in SPSS.
  • Go to Analyze > Descriptive Statistics > Descriptives.
  • Select the variables you want to analyze (e.g., a variable measuring how environmentally friendly someone is on a scale of 1 to 7).
  • Click 'OK' to generate the descriptive statistics.

• Interpreting Descriptive Statistics:

  • Example: If the average on a 1-7 scale is 3.9, it indicates responses are close to the middle (4), suggesting no strong agreement or disagreement.
  • Mean centering: Transforming the scale so that the middle point (4) becomes zero, allowing for positive and negative deviations.
  • Remember: The scale ranges from strongly disagree(1) to strongly agree(7).

Options for Basic Descriptives

• Available Options

  • Range of the data.
  • Minimum and maximum values.
  • Standard deviation.
  • Display order (ascending, descending).
  • Bootstrapping (not necessary for the assignment).

• Pasting Syntax

  • Paste the syntax for saving and later use.

• Frequency Variables

  • Obtain frequencies for all variables in the dataset.
  • Output can be saved to a file (e.g., output space/file='blah blah blah blah.sspv').

• Running All Commands

  • SPSS allows running all commands at once.
  • Output may include nonsensical results for certain variables (e.g., respondent ID).

• SPSS Interface

  • The left-hand side of the SPSS output displays the actions performed.
  • Clicking on items allows navigation within the output.

• Saving Output

  • SPSS automatically saves output if specified in the syntax.

• Using Output

  • The output file contains all the performed operations and results.

Response Rates and Incentives

• Analyzing Response Rates

  • Crucial to examine response rates, especially in professional research, to ensure appropriate incentives.
  • Example: If 80% reject participation, the incentive may be inadequate.

• Informed Consent

  • Poorly written informed consent can lead to high dropout rates.
  • High punch-out rates (participants stopping midway) indicate issues with the survey design or informed consent.

• Copying Output to Word

  • Paste output into Word as a picture for cleaner formatting and better aspect ratio.
  • Pasting as a picture maintains the visual integrity compared to other methods.
  • Advantage: easy to incorporate graphs and tables into reports.
  • SPSS output remains available for further analysis if needed.

Frequency Tables

• Generating Frequency Tables

  • Analyze > Descriptive Statistics > Frequencies.
  • Frequency tables show the distribution of responses (e.g., 14% strongly agreed).
  • Display both labels and levels for clarity.
  • Measures of central tendency: mean, median, mode.

• Measures of Central Tendency

  • Mode: The most common value.
  • Median: The value at which 50% of responses fall below and 50% fall above.
  • Mean: The mathematical average of the values.

• Sorting Frequency Tables

  • Sort by count to easily identify the mode.
  • Analyze > Descriptive > Frequencies > Format > Descending Counts.

• Advantages: Frequency tables provide counts for each response.

• Additional Statistics: Frequency tables can include additional statistics.

Charts

• Frequency tables can include charts (e.g., bar charts).

• Sorting for Charts: Sort data appropriately for meaningful chart representation.

• Pasting Charts: Charts can be pasted into documents.

Explore Function

• Deep Dive Analysis: Analyze > Explore for in-depth data exploration.

• Cross-Tabulation: Allows cross-tabulation, such as examining age by responses to daily life questions.

• Output: Provides extensive statistics, including means and breakdowns by categories.

• Box Plots: Useful for visualizing data distribution.

• Usage: Useful for a deeper understanding of variable relationships.

Charts and Graphs

• Charts and Graphs: Charts can be created through Analyze > Charts.

• Legacy Dialogs: SPSS provides legacy chart options.

• Chart Builder: Newer versions include a chart builder with various chart types (e.g., pie charts).

• Customization: Chart Builder allows customization of chart elements and colors.

Graphics in SPSS

• Keep output SPSS output file (.spv) because it contains all manipulations. Then copy into Word as picture.

• File Size: Graphics increase file size significantly.

• Aspect Ratio: Graphics maintain aspect ratio better when pasted as a picture.

• Manipulation: Perform all manipulations within SPSS if following the copy-as-picture approach.

• Excel for charts: Excel can also be used to create charts and graphs.

• Paste Special: Paste as text into excel to provide the most flexibility.

Additional Points

• There is no right or wrong way. Whatever you like, do it. The most important thing is to answer the questions asked for the projects.

• Links for practice datasets may not work due to cloud and VPN changes. Check IBM for test datasets and play around with them.

FAQs and Data Cleaning

• Data Export Issues

  • Often due to using the free version instead of the approved UPS version.

• Data Cleaning

  • Data cleaning: Is the majority of real-life analytics work.
  • Always clean your data to ensure accuracy.
  • Perfectly clean data is rare.

• Missing Data

  • Consider the ratio of answered to unanswered questions.
  • If a respondent answers the majority of questions, keep their data.
  • Code missing data appropriately (e.g., -99 for seen but unanswered).
  • Analyze patterns of missing data to identify potential survey design flaws.

• Blank Responses

  • Assess the ratio of answered to unanswered questions.
  • If a respondent omits a significant portion (e.g., >50%), consider omitting their data.
  • Incomplete engagement may indicate a lack of care in answering honestly.

  ratio = \frac{answered \, questions}{total \, questions}