Presentation is loading. Please wait.

Presentation is loading. Please wait.

The Next Generation of Access to NCES Data on Postsecondary, Adult, and Career Education Matthew Soldner Postsecondary Longitudinal and Sample Studies.

Similar presentations


Presentation on theme: "The Next Generation of Access to NCES Data on Postsecondary, Adult, and Career Education Matthew Soldner Postsecondary Longitudinal and Sample Studies."— Presentation transcript:

1 The Next Generation of Access to NCES Data on Postsecondary, Adult, and Career Education Matthew Soldner Postsecondary Longitudinal and Sample Studies Program National Center for Education Statistics April 8, 2010

2 W HAT IS P OWER S TATS ? Your portal to a series of NCES sample surveys. A way to generate tables of national estimates. A way to conduct correlational analyses, including weighted least squares and logistic regressions.

3 NCES P OST S ECONDARY D ATA S OURCES Universe data is not in PowerStats IPEDS Sample data is in PowerStats National Study of Postsecondary Faculty (NSOPF) National Postsecondary Student Aid Study (NPSAS) A cross-section of undergraduate and graduate students at all Title IV institutions Beginning Postsecondary Students Study (BPS) A longitudinal NPSAS subsample focused on first-time beginning students Baccalaureate and Beyond (B&B) A longitudinal NPSAS subsample focused on baccalaureate recipients

4 T ABLES OF N ATIONAL E STIMATES Generates percentage of population in each category of a variable, displayed in columns. Example: Distribution of dependency status by age Computes averages, medians, or percentages for your selected variables, displayed in columns. Example: Average Pell Grant by Parental Income Group Produces values of a continuous variable at centiles, displayed in columns. Example: Distribution of Grade Point Average by Major

5 C ORRELATIONAL A NALYSES Estimate linear relationship between predictor variables and a continuous outcome. Example: Publications, based on faculty and institutional characteristics Estimate relationship between predictor variables and a binary outcome. Example: Student persistence, based on student and institutional characteristics Compute correlations between two or more variables. Example: Relationship among student aid variables

6 W HO A RE P OWER S TATS U SERS ? Policy and research analysts focused on postsecondary students. Research faculty. Each and every one of you!

7 M E ? W HY W OULD I U SE P OWER S TATS ? “But I came here because of IPEDS, and IPEDS isn’t going to be a part of PowerStats, right?”

8 M E ? W HY W OULD I U SE P OWER S TATS ? We think you’ll find that PowerStats will become your source for context-setting, nationally-representative data about the student experience.

9 M E ? W HY W OULD I U SE P OWER S TATS ? We think you’ll find that PowerStats will become your source for context-setting, nationally-representative data about the student experience.

10 H OW TO A CCESS AND U SE P OWER S TATS

11 T HREE R ESEARCH Q UESTIONS 1. Your provost is interested in how the distribution of student participation in various majors (as opposed to completions in those majors) match up to national estimates.

12 T HREE R ESEARCH Q UESTIONS 2. Your campus has been intentionally focused on the experience of first-time, beginning students with dependents, and have done a survey of student financing including their aid, their work habits, and their daycare costs. You want to see if campus results mirror national estimates.

13 T HREE R ESEARCH Q UESTIONS 3. You’ve been doing some persistence modeling using information from your campus data warehouse. You want to see how your results fit with those at a national level.

14 D ISTRIBUTION OF M AJORS If we’re talking about the distribution of majors for all students, then we’ll be using NPSAS, our nationally-representative study. Results in five (and a half) simple steps! 1. Choosing our dataset. 2. Identifying “percentage distribution” as the correct table type. 3. Locating the desired row and column variables. 4. Applying any filters. 5. Running the table.

15

16

17

18

19

20

21

22

23 F INANCING D ATA FOR S TUDENTS W ITH D EPENDENTS Since we’re interested in first-time, beginning students, we are using the Beginning Postsecondary Students Study. Results in five simple steps! 1. Choosing our dataset. 2. Identifying “averages” as the correct table type. 3. Locating the desired row and column variables. Remember: COLUMNS hold our averages … 4. Applying the proper filters to select only those students with dependents. 5. Running the table. What? You want to know standard errors or confidence intervals? Fine.

24

25

26

27

28 P ERSISTENCE R EGRESSIONS Since we’re interested in the persistence of first- time, beginning students, we are using a longitudinal component of the Beginning Postsecondary Students Study. Results in five simple steps! 1. Choosing our dataset. 2. Choosing our weight. 3. Identifying “logistic” as the correct regression type. 4. Locating the desired dependent and independent variables. 5. Running the regression.

29

30

31

32

33

34 THANK YOU! What Questions Do You Have for Me? Want to learn more about: NPSAS?nces.ed.gov/surveys/npsas, or Tracy Hunt-White BPS?nces.ed.gov/surveys/bps, or Matthew Soldner B&B or PowerStats? nces.ed.gov/surveys/b&b, or Ted Socha NSOPF or our program? nces.ed.gov/surveys/nsopf, or Linda Zimbler


Download ppt "The Next Generation of Access to NCES Data on Postsecondary, Adult, and Career Education Matthew Soldner Postsecondary Longitudinal and Sample Studies."

Similar presentations


Ads by Google