The independent news organization of Duke University

Chronicle to release survey data on Class of 2024: Inside the methodology and limitations

Editor's note: This story is part of a series about the Class of 2024 based on a survey conducted by The Chronicle. You can see all of our survey coverage here.

Interactives by Selena Qian. 

This week, The Chronicle will release survey data about the Class of 2024.

For the fourth consecutive year, we surveyed the first-year class about their lifestyles, demographics, plans at Duke and more. Questions ranged from high school test scores and religious beliefs to their approximate family income and intention to rush.

Many of the questions were similar to last year’s survey. This year, given the presidential election and COVID-19 pandemic, additional questions included who students voted for, where students voted, whether they were able to find community and whether household members lost their jobs or were furloughed during the pandemic. 

Questions about tenting and basketball were removed given the uncertainty of tenting and in-person basketball games this year. The survey also removed questions about cell phones, operating systems and social media use to make room for the additional voting and COVID-19 questions.

In total, this year’s survey asked 67 questions, compared to 35 questions last year. 

The results of this survey will be released throughout the coming week in a series of stories and will give a deeper look at Duke’s newest undergraduate students.

The survey was administered from Oct. 27 to Nov. 17, and 299 first-years filled out 99%-100% complete responses, with 442 total students starting a response. In total, 18.8% of the first-year class completed the survey, and three random prizes were awarded to participants. 

Professor of Statistical Science Jerry Reiter previously told The Chronicle about the potential limitations of the survey. He said that students’ reasons for opting out of the survey—including lack of access to the survey or low enthusiasm for Duke—could make the results less representative of the class.

Since this year’s survey asked more than 60 questions, students may have not chosen to complete the questionnaire due to time constraints. Reiter wrote in an email that this could result in the nonrespondents being systematically different than the respondents, which could skew the results of the survey. 

Below, we compare our survey data with the official Class of 2024 profile from the Office of Undergraduate Admissions and additional data sent to The Chronicle by Dean of Undergraduate Admissions Christoph Guttentag. This shows how our survey responses may differ from the overall class.

Academics

The Chronicle’s survey data slightly overrepresents the number of students who attended a public high school and the number that attended a private high school. However, the data is not fully comparable since official data further delineates students who were schooled outside of the United States, which The Chronicle’s survey does not.

Our data also slightly overrepresents the number of Regular Decision applicants in the Class of 2024. While 52.2% of our respondents applied Early Decision to Duke, 49.4% of the actual class did. 

Our survey data slightly overrepresents the number of students enrolled in the Pratt School of Engineering. While 20.4% of the first-year class is in Pratt and 79.6% is in the Trinity College of Arts and Sciences, 23.1% of our respondents are in Pratt and 76.9% are in Trinity.


The Chronicle’s survey also overrepresents the number of FOCUS students. 42.5% of our respondents were FOCUS participants, while an estimated 20-25% of first-years partake in the program each year, according to the FOCUS website.

Race and ethnicity

The Chronicle’s survey gathered race and ethnicity data differently than Duke’s admissions office by characterizing Hispanic and Latino as an ethnicity rather than a race, including an “other” option and allowing multiracial students to select multiple races. Therefore, our data and the official data on race and ethnicity data are not fully comparable. 

Our survey asked respondents first whether they were of Hispanic/Latino/Spanish origin, then asked which racial group(s) they most identified with. Of the respondents, 13.1% are Hispanic or Latinx. Of non-Hispanic respondents, 44.8% are white, 38.2% are Asian, 9.7% are multiracial, 5.8% are Black or African American, 1.2% marked “Other” and 0.39% are American Indian or Alaskan Native. 

Official diversity data records 41% white students, 30% Asian students, 11% Black or African American students, 10% Hispanic and Latinx students and 2% Native American or Alaskan Native students, with 5% of students who did not indicate race. 

Guttentag confirmed in an email to The Chronicle that the University’s official racial diversity data did not include an “other” category and did not give multiracial students the option to select multiple races. 

He added that he hopes in future profiles Duke can “capture students who identify as being of more than one race more accurately.”


Geographic and socioeconomic diversity

The New York Times published data about Duke’s socioeconomic diversity based on tax records in 2016. Although there are differences between The Chronicle’s and the Times’ analyses in the metrics used to report the data, our survey data is generally consistent with the Times’ data. The Times’ data shows that the median family income of Duke students is $186,700, and the median family income bracket from our data was between $125,000 - $250,000. 

The Chronicle’s survey data was comparable to the official number of students receiving financial aid. According to Duke data, 50% of the Class of 2024 receive financial aid, compared to 51% of our respondents.

Additionally, 4.3% of our respondents—compared to 9% of the Class of 2024—are international students, indicating that our survey underrepresents the percentage of international students and overrepresents the proportion of domestic students.

Other identity groups

Whereas 49% of the Class of 2024 are female-identifying and 51% are male-identifying, 60.7% of our respondents identified as cisgender women, 36.2% identified as cisgender men, 1.7% identified as genderqueer or nonbinary, 1% identified as other and 0.34% identified as agender. Though the Chronicle included more options for gender identity than Duke’s data, the discrepancy indicates that our data overrepresents the number of female-identifying students and underrepresents the number of male-identifying students. 

Our survey data aligned closely with the number of first-generation college students. Whereas 9% of all first-years are first-generation students, 9.1% of our respondents are and 1% preferred not to say.

The Chronicle’s survey slightly overrepresented the number of legacy students. Compared to 11% of the Class of 2024, 14.4% of our respondents are legacy students. 

The survey responses also underrepresent the number of recruited varsity athletes at Duke. Whereas 8.9% of the Class of 2024 are recruited varsity athletes, only 2% of our respondents are.

Discussion

Share and discuss “Chronicle to release survey data on Class of 2024: Inside the methodology and limitations” on social media.