Difference between revisions of "National Origin or National Location"

From Penn Center for Learning Analytics Wiki
Jump to navigation Jump to search
(terminology corrections)
m (Add empty lines)
 
(3 intermediate revisions by the same user not shown)
Line 1: Line 1:
Švábenský et al. (2024) [https://educationaldatamining.org/edm2024/proceedings/2024.EDM-posters.82/2024.EDM-posters.82.pdf pdf]


Ogan et al. (2015) [https://link.springer.com/content/pdf/10.1007/s40593-014-0034-8.pdf pdf]
*Classification models for predicting grades (worse than an average grade, “unsuccessful”, or equal/better than an average grade, “successful”)
*Investigating bias based on university students' regional background in the context of the Philippines
*Demographic groups based on 1 of 5 locations from which students accessed online courses in Canvas
*Bias evaluation using AUC, weighted F1-score, and MADD showed consistent results across all groups, no unfairness was observed


* Multi-national models predicting learning gains from student's help-seeking behavior
* Models built on only U.S. or combined data sets performed extremely poorly for Costa Rica
* Models performed better when built on and applied for the same country, except for Philippines where model built on that country which was outperformed slightly by model built on U.S. data


Li et al. (2021) [https://arxiv.org/pdf/2103.15212.pdf pdf]


*Model predicting student achievement on the standardized examination PISA
*Inaccuracy of the U.S.-trained model was greater for students from countries with lower scores of national development (e.g. Indonesia, Vietnam, Moldova)


Li et al. (2021) [https://arxiv.org/pdf/2103.15212.pdf pdf]


* Model predicting student achievement on the standardized examination PISA
Wang et al. (2018) [https://www.researchgate.net/publication/336009443_Monitoring_the_performance_of_human_and_automated_scores_for_spoken_responses pdf]
* Inaccuracy of the U.S.-trained model was greater for students from countries with lower scores of national development (e.g. Indonesia, Vietnam, Moldova)


*Automated scoring model for evaluating English spoken responses
*SpeechRater gave a significantly lower score than human raters for German students
*SpeechRater scored gave higher scores than human raters for Chinese students, with H1-rater scores higher than mean


Wang et al. (2018) [https://www.researchgate.net/publication/336009443_Monitoring_the_performance_of_human_and_automated_scores_for_spoken_responses pdf]


* Automated scoring model for evaluating English spoken responses
Ogan et al. (2015) [https://link.springer.com/content/pdf/10.1007/s40593-014-0034-8.pdf pdf]
* SpeechRater gave a significantly lower score than human raters for German students
* SpeechRater scored gave higher scores than human raters for Chinese students, with H1-rater scores higher than mean


*Multi-national models predicting learning gains from student's help-seeking behavior
*Models built on only U.S. or combined data sets performed extremely poorly for Costa Rica
*Models performed better when built on and applied for the same country, except for Philippines where model built on that country which was outperformed slightly by model built on U.S. data


Bridgeman et al. (2009) [https://www.researchgate.net/publication/242203403_Considering_Fairness_and_Validity_in_Evaluating_Automated_Scoring page]


* Automated scoring models for evaluating English essays, or e-rater
Bridgeman et al. (2012) [https://www.tandfonline.com/doi/pdf/10.1080/08957347.2012.635502?needAccess=true pdf]


* E-Rater gave significantly better scores than human rater for TOEFL essays (independent task) written by speakers of Chinese and Korean
*A later version of automated scoring models for evaluating English essays, or e-rater
* E-Rater correlated poorly with human rater and gave better scores than human rater for GRE essays (both issue and argument prompts) written by Chinese speakers
*E-rater gave better scores for test-takers from Chinese speakers (Mainland China, Taiwan, Hong Kong) and Korean speakers when assessing TOEFL (independent prompt) essay
*E-rater gave lower scores for Arabic, Hindi, and Spanish speakers when assessing their written responses to independent prompt in TOEFL




Bridgeman et al. (2009) [https://www.researchgate.net/publication/242203403_Considering_Fairness_and_Validity_in_Evaluating_Automated_Scoring page]


Bridgeman et al. (2012) [https://www.tandfonline.com/doi/pdf/10.1080/08957347.2012.635502?needAccess=true pdf]
*Automated scoring models for evaluating English essays, or e-rater
 
*E-Rater gave significantly better scores than human rater for TOEFL essays (independent task) written by speakers of Chinese and Korean
* A later version of automated scoring models for evaluating English essays, or e-rater
*E-Rater correlated poorly with human rater and gave better scores than human rater for GRE essays (both issue and argument prompts) written by Chinese speakers
* E-rater gave better scores for test-takers from Chinese speakers (Mainland China, Taiwan, Hong Kong) and Korean speakers when assessing TOEFL (independent prompt) essay
* E-rater gave lower scores for Arabic, Hindi, and Spanish speakers when assessing their written responses to independent prompt in TOEFL

Latest revision as of 20:13, 1 September 2024

Švábenský et al. (2024) pdf

  • Classification models for predicting grades (worse than an average grade, “unsuccessful”, or equal/better than an average grade, “successful”)
  • Investigating bias based on university students' regional background in the context of the Philippines
  • Demographic groups based on 1 of 5 locations from which students accessed online courses in Canvas
  • Bias evaluation using AUC, weighted F1-score, and MADD showed consistent results across all groups, no unfairness was observed


Li et al. (2021) pdf

  • Model predicting student achievement on the standardized examination PISA
  • Inaccuracy of the U.S.-trained model was greater for students from countries with lower scores of national development (e.g. Indonesia, Vietnam, Moldova)


Wang et al. (2018) pdf

  • Automated scoring model for evaluating English spoken responses
  • SpeechRater gave a significantly lower score than human raters for German students
  • SpeechRater scored gave higher scores than human raters for Chinese students, with H1-rater scores higher than mean


Ogan et al. (2015) pdf

  • Multi-national models predicting learning gains from student's help-seeking behavior
  • Models built on only U.S. or combined data sets performed extremely poorly for Costa Rica
  • Models performed better when built on and applied for the same country, except for Philippines where model built on that country which was outperformed slightly by model built on U.S. data


Bridgeman et al. (2012) pdf

  • A later version of automated scoring models for evaluating English essays, or e-rater
  • E-rater gave better scores for test-takers from Chinese speakers (Mainland China, Taiwan, Hong Kong) and Korean speakers when assessing TOEFL (independent prompt) essay
  • E-rater gave lower scores for Arabic, Hindi, and Spanish speakers when assessing their written responses to independent prompt in TOEFL


Bridgeman et al. (2009) page

  • Automated scoring models for evaluating English essays, or e-rater
  • E-Rater gave significantly better scores than human rater for TOEFL essays (independent task) written by speakers of Chinese and Korean
  • E-Rater correlated poorly with human rater and gave better scores than human rater for GRE essays (both issue and argument prompts) written by Chinese speakers