How risky are one-shot cross-sectional assessments of undergraduate students?

Alexander W. Astin, Jenny J Lee

Research output: Contribution to journalArticle

47 Citations (Scopus)

Abstract

The purpose of this study was to explore the extent of possible input bias associated with the rapidly growing use by colleges and universities of one-time cross-sectional assessments of students. The results presented in this study demonstrate that cross-sectional assessments of enrolled college students are very difficult to interpret because they inevitably reflect characteristics of the same students when they first entered college. For those forms of engagement that have to do with how students allocate their time, it appears that the majority of variation among institutions is attributable to entering freshman characteristics rather than to institutional policies or practices. Thus, rather than relying solely on student outcome data, institutions should ideally also collect pretest or input information from the same students when they first matriculate.

Original languageEnglish (US)
Pages (from-to)657-672
Number of pages16
JournalResearch in Higher Education
Volume44
Issue number6
DOIs
StatePublished - Dec 2003
Externally publishedYes

Fingerprint

student
university
trend
time

Keywords

  • Assessment
  • Cross-sectional
  • Inferential risk
  • Longitudinal
  • Outcomes

ASJC Scopus subject areas

  • Education

Cite this

How risky are one-shot cross-sectional assessments of undergraduate students? / Astin, Alexander W.; Lee, Jenny J.

In: Research in Higher Education, Vol. 44, No. 6, 12.2003, p. 657-672.

Research output: Contribution to journalArticle

@article{ca68cbe0302049b187a5a239f1bc7dcc,
title = "How risky are one-shot cross-sectional assessments of undergraduate students?",
abstract = "The purpose of this study was to explore the extent of possible input bias associated with the rapidly growing use by colleges and universities of one-time cross-sectional assessments of students. The results presented in this study demonstrate that cross-sectional assessments of enrolled college students are very difficult to interpret because they inevitably reflect characteristics of the same students when they first entered college. For those forms of engagement that have to do with how students allocate their time, it appears that the majority of variation among institutions is attributable to entering freshman characteristics rather than to institutional policies or practices. Thus, rather than relying solely on student outcome data, institutions should ideally also collect pretest or input information from the same students when they first matriculate.",
keywords = "Assessment, Cross-sectional, Inferential risk, Longitudinal, Outcomes",
author = "Astin, {Alexander W.} and Lee, {Jenny J}",
year = "2003",
month = "12",
doi = "10.1023/A:1026175525173",
language = "English (US)",
volume = "44",
pages = "657--672",
journal = "Research in Higher Education",
issn = "0361-0365",
publisher = "Springer Netherlands",
number = "6",

}

TY - JOUR

T1 - How risky are one-shot cross-sectional assessments of undergraduate students?

AU - Astin, Alexander W.

AU - Lee, Jenny J

PY - 2003/12

Y1 - 2003/12

N2 - The purpose of this study was to explore the extent of possible input bias associated with the rapidly growing use by colleges and universities of one-time cross-sectional assessments of students. The results presented in this study demonstrate that cross-sectional assessments of enrolled college students are very difficult to interpret because they inevitably reflect characteristics of the same students when they first entered college. For those forms of engagement that have to do with how students allocate their time, it appears that the majority of variation among institutions is attributable to entering freshman characteristics rather than to institutional policies or practices. Thus, rather than relying solely on student outcome data, institutions should ideally also collect pretest or input information from the same students when they first matriculate.

AB - The purpose of this study was to explore the extent of possible input bias associated with the rapidly growing use by colleges and universities of one-time cross-sectional assessments of students. The results presented in this study demonstrate that cross-sectional assessments of enrolled college students are very difficult to interpret because they inevitably reflect characteristics of the same students when they first entered college. For those forms of engagement that have to do with how students allocate their time, it appears that the majority of variation among institutions is attributable to entering freshman characteristics rather than to institutional policies or practices. Thus, rather than relying solely on student outcome data, institutions should ideally also collect pretest or input information from the same students when they first matriculate.

KW - Assessment

KW - Cross-sectional

KW - Inferential risk

KW - Longitudinal

KW - Outcomes

UR - http://www.scopus.com/inward/record.url?scp=0345863470&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0345863470&partnerID=8YFLogxK

U2 - 10.1023/A:1026175525173

DO - 10.1023/A:1026175525173

M3 - Article

AN - SCOPUS:0345863470

VL - 44

SP - 657

EP - 672

JO - Research in Higher Education

JF - Research in Higher Education

SN - 0361-0365

IS - 6

ER -