1 / 30

13. Analysis Demonstration: Descriptive/Comparative Analysis Using Longitudinal Data

13. Analysis Demonstration: Descriptive/Comparative Analysis Using Longitudinal Data. Prerequisites. Recommended modules to complete before viewing this module 1. Introduction to the NLTS2 Training Modules 2. NLTS2 Study Overview 3. NLTS2 Study Design and Sampling NLTS2 Data Sources, either

yuki
Download Presentation

13. Analysis Demonstration: Descriptive/Comparative Analysis Using Longitudinal Data

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. 13.Analysis Demonstration: Descriptive/Comparative AnalysisUsing Longitudinal Data

  2. Prerequisites • Recommended modules to complete before viewing this module • 1. Introduction to the NLTS2 Training Modules • 2. NLTS2 Study Overview • 3. NLTS2 Study Design and Sampling • NLTS2 Data Sources, either • 4. Parent and Youth Surveys or • 5. School Surveys, Student Assessments, and Transcripts • 9. Weighting and Weighted Standard Errors

  3. Prerequisites • Recommended modules to complete before viewing this module (cont’d) • NLTS2 Documentation • 10. Overview • 11. Data Dictionaries • 12. Quick References • Helpful • Implications for analysis • 6. Data Content • 7. Parent/Youth Survey Data

  4. Overview • Posing questions to the NLTS2 database • Types of questions; getting to the answers • Comparative variables • Analysis demonstration using longitudinal data • Analysis plan for this demonstration • Results output for this demonstration • Interpretation of results for this demonstration • Review steps required for this demonstration • Closing • Important information

  5. Posing Questions to the NLTS2database • Types of questions • Descriptive questions—e.g., means, frequency distributions for an identified group • Comparative questions—e.g., differences between subgroups of interest, tests of significance • Longitudinal questions—i.e., relationships between variables measured at different time points • Explanatory questions—i.e., questions regarding relationships between variables, usually requiring controls for covariates

  6. Examples of types of questions • Descriptive/comparative • How proficient are high school students with disabilities in reading? • How does proficiency differ by disability category? By race/ethnicity? • Longitudinal/comparative • If youth has been employed, how does income change over time for youth with disabilities as a group? How does it change at the individual level? • How do changes in wages over time differ by disability category?By race/ethnicity? • Explanatory • What factors relate to variations in the employment of youth with disabilities?

  7. Getting to the answer • Select • Analysis approach from options appropriate to the question. • Software procedure(s) to execute approach. • Execute plan. • Develop an analysis plan. • Identify/create variables. • Identify data source. • Specify variable names. • Be aware of variable forms (e.g., continuous, categorical,) and respondent subset. • Restructure variables if applicable (e.g., regroup into different categories). • Create variables as needed (e.g., scales from multiple items).

  8. Comparative variables • A set of demographic variables are commonly used for comparative analyses in NLTS2 reports and publications. • NLTS2 demographic variables are by wave and included in every school and parent/youth sourced files. • The demographic variables are • Primary disability category • Age category • Race/ethnicity • Gender • Household income category

  9. Comparative variables • Grade level and urbanicity are used with school-based data. • Demographic variables are based on location of the school district where youth attended secondary school. • School related demographic variables are included with school data files but not applicable to parent/youth survey files.

  10. Analysis demonstration usinglongitudinal data • Research questions for this demonstration • How do wages change over time for youth with disabilities as a group? • How do they change at the individual level? • How do changes in youth wages over time differ by parent’s household income?

  11. Analysis plan for this demonstration • Select procedures that will describe these changes. • In this demonstration, we will use procedures that provide weighted frequencies, percentages, means, and standard errors. • Select procedures to make comparisons. • In this demonstration, we will produce crosstabulations and means by a comparison variable and determine whether differences are statistically significant.

  12. Analysis plan for this demonstration • Select variables that measure change. • Compare items in Parent/Youth Wave 2 with Parent/Youth Wave 5. • Wage youth earned at current or most recent job (Wave 2 np2HourlyWage compared with Wave 5 np5T4h_L4h). • Select comparison variables. • Comparison variables are often demographics such as primary disability category, gender, and race/ethnicity. • Parent/guardian’s household income will be the comparison variable in this demonstration (W5_IncomeHdr2009).

  13. Analysis plan for this demonstration • Compare time 1 and time 2. • One approach to observe a difference between time 1 and time 2 is to compare a time 1 mean with a time 2 mean calculated across the entire sample. • To measure individual difference, an approach is to calculate the difference between that youth’s time 1 measure and his or her time 2 measure. • To calculate the difference between time 1 and time 2, there must be a value for both time 1 and time 2 to be included in this analysis.

  14. Analysis plan for this demonstration • Missing values • Expect to have fewer cases to analyze in a longitudinal analysis than in a separate analysis of either time 1 or time 2. • Youth must have data in both waves to be included. • Create new variables • Difference = time 2 – time1 • The order is important; subtracting time 1 from time 2 will result in a positive number if the change has been in a positive direction.

  15. Results output for this demonstration • Means: hourly wages in Wave 2 and Wave 5 • Mean: difference in wages between Wave 2 and Wave 5 • Percentage who had a decrease or increase in wages • Crosstabs by parent/guardian household income

  16. NLTS2 restricted-use data NLTS2 data are restricted. Data used in these presentations are from a randomly selected subset of the restricted-use NLTS2 data. Results in these presentations cannot be replicated with the NLTS2 data licensed by NCES.

  17. Means of Hourly Wages in Wave 2 and Wave 5 Results output for this demonstration These results cannot be replicated with full dataset; all outputin modules generated with a random subset of the full data.

  18. Mean Difference in Wages between Wave 2 and Wave 5 Results output for this demonstration These results cannot be replicated with full dataset; all outputin modules generated with a random subset of the full data.

  19. Percentage Who Had a Decrease or Increase in Wages between Waves 2 and 5 Results output for this demonstration These results cannot be replicated with full dataset; all outputin modules generated with a random subset of the full data.

  20. Mean Difference in Wages, by Household Income between Wave 2 and Wave 5 Results output for this demonstration These results cannot be replicated with full dataset; all outputin modules generated with a random subset of the full data.

  21. Percentage Decrease or Increase in Wages by Household Income between Waves 2 and 5 Results output for this demonstration These results cannot be replicated with full dataset; all outputin modules generated with a random subset of the full data.

  22. Interpretation of results for thisdemonstration These results cannot be replicated with full dataset; all outputin modules generated with a random subset of the full data. • Is it significant? • We use an Excel worksheet. • Enter (1) the percentage and standard error for a category and (2) the percentage and standard error for the comparison category. • The significance is calculated by a formula stored in Excel.

  23. Interpretation of results for this demonstration • Interpreting longitudinal results. • A positive number indicates a positive change between time 1 and time 2 for difference in wage. • A negative number indicates a negative change between time 1 and time 2. • Numbers close to 0 indicate little or no change. • What does it mean? • A positive or negative change may be simple to interpret, but what does little or no change mean? • What explains the change? • For example, postsecondary attendance might mean a change in employment.

  24. Interpretation of results for this demonstration • Was anything significant ? • Not with the randomly selected subset of data. • The full data set might have a different result if the standard errors were smaller. • Or perhaps household income was not the best comparison variable to use. • Things to think about • What other comparison variables might show significant differences between comparison groups? • Are there other difference in employment measures that might be worth exploring? • Were these the best waves of data to compare? • Should any types of youth be included or excluded from this analysis?

  25. Interpretation of results for this demonstration • More things to think about • Did we have a good comparison group? • Would it have been better to select Waves 3 and 5 instead of 2 and 5 as more youth are out of secondary school in Wave 3 than in 2? • Would it have been better to run the analysis on a subgroup of only those who were out of secondary school at each wave? • The answer is…it is important to select the appropriate subsample as well as to select the appropriate variables.

  26. Review steps required for thisdemonstration • Plan for analysis • Identify the research question. • Find the variables needed. • Determine if any variables need to recoded. • Collapse categories. • Create a longitudinal measure. • Identify the appropriate subsample if applicable.

  27. Review steps required for thisdemonstration • Plan for analysis (cont’d) • Identify the appropriate procedures. • Descriptive • Crosstabs • Means • Means by comparison variables if calculating means • Test significance of comparisons

  28. Closing • Topics discussed in this module • Posing questions to the NLTS2 database • Comparative variables • Analysis demonstration using longitudinal data • Analysis plan for this demonstration • Results output for this demonstration • Interpretation of results for this demonstration • Review steps required for this demonstration

  29. Closing • Next module • 14a. Accessing Data Files in SPSS or • 14b. Accessing Data Files in SAS

  30. Important information • NLTS2 website contains reports, data tables, and other project-related information http://nlts2.org/ • Information about obtaining the NLTS2 database and documentation can be found on the NCES website http://nces.ed.gov/statprog/rudman/ • General information about restricted data licenses can be found on the NCES websitehttp://nces.ed.gov/statprog/instruct.asp • E-mail address: nlts2@sri.com

More Related