ABSTRACT
We evaluate statistical inference procedures for small-scale IR experiments that involve multiple comparisons against the baseline. These procedures adjust for multiple comparisons by ensuring that the probability of observing at least one false positive in the experiment is below a given threshold. We use only publicly available test collections and make our software available for download. In particular, we employ the TREC runs and runs constructed from the Microsoft learning-to-rank (MSLR) data set. Our focus is on non-parametric statistical procedures that include the Holm-Bonferroni adjustment of the permutation test p-values, the MaxT permutation test, and the permutation-based closed testing. In TREC-based simulations, these procedures retain from 66% to 92% of individually significant results (i.e., those obtained without taking other comparisons into account). Similar retention rates are observed in the MSLR simulations. For the largest evaluated query set size (i.e., 6400), procedures that adjust for multiplicity find at most 5% fewer true differences compared to unadjusted tests. At the same time, unadjusted tests produce many more false positives.
- Anonymous. Guidance for Industry - E9 Statistical Principles for Clinical Trials. Technical report, U.S. Department of Health and Human Services - Food and Drug Administration, Center for Drug Evaluation and Research, Center for Biologics Evaluation and Research, ICH, 1998.Google Scholar
- R. Bender and S. Lange. Adjusting for multiple testing--when and how? Journal of Clinical Epidemiology, 54(4):343--349, 2001.Google ScholarCross Ref
- Y. Benjamini and Y. Hochberg. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society. Series B (Methodological), 57(1):289--300, 1995.Google ScholarCross Ref
- R. Blanco and H. Zaragoza. Beware of relatively large but meaningless improvements. Technical report YL-2011-001, Yahoo! Research, 2011.Google Scholar
- C. Buckley, D. Dimmick, I. Soboroff, and E. Voorhees. Bias and the limits of pooling for large collections. Information Retrieval, 10:491--508, 2007. Google ScholarDigital Library
- R. J. Cabin and R. J. Mitchell. Bonferroni or not Bonferroni: when and how are the questions. Bulletin of the Ecological Society of America, 81(3):246--248, 2000.Google Scholar
- B. A. Carterette. Multiple testing in statistical analysis of systems-based information retrieval experiments. ACM Trans. Inf. Syst., 30(1):4:1--4:34, Mar. 2012. Google ScholarDigital Library
- O. Chapelle, D. Metlzer, Y. Zhang, and P. Grinspan. Expected reciprocal rank for graded relevance. In Proceeding of the 18th ACM conference on Information and knowledge management, CIKM '09, pages 621--630, New York, NY, USA, 2009. ACM. Google ScholarDigital Library
- C. L. A. Clarke, N. Craswel, I. Soboroff, and G. V. Cormack. Overview of TREC 2010 Web track. InmboxTREC-19: Proceedings of the Nineteenth Text REtrieval Conference, 2010.Google Scholar
- G. V. Cormack and T. R. Lynam. Validity and power of t-test for comparing map and gmap. In Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval, SIGIR '07, pages 753--754, New York, NY, USA, 2007. ACM. Google ScholarDigital Library
- S. Dudoit, J. Schaffer, and J. Boldrick. Multiple hypothesis testing in microarray experiments. Statistical Science, 18(1):71--103, 2003.Google ScholarCross Ref
- B. Efron and R. Tibshirani. An Introduction to the Bootstrap. Monographs on Statistics and Applied Probability. Chapman & Hall, 1993.Google ScholarCross Ref
- S. Holm. A Simple Sequentially Rejective Multiple Test Procedure. Scandinavian Journal of Statistics, 6:65--70, 1979.Google Scholar
- Y. Huang, H. Xu, V. Calian, and J. C. Hsu. To permute or not to permute. Bioinformatics, 22(18):2244--2248, 2006. Google ScholarDigital Library
- E. L. Lehmann and J. P. Romano. Generalizations of the familywise error rate. Annals of Statistics, 33(3):1138--1154, 2005.Google ScholarCross Ref
- R. Marcus, P. Eric, and K. R. Gabriel. On closed testing procedures with special reference to ordered analysis of variance. Biometrika, 63(3):655--660, 1976.Google ScholarCross Ref
- M. Matsumoto and T. Nishimura. Mersenne twister: a 623-dimensionally equidistributed uniform pseudo-random number generator. ACM Trans. Model. Comput. Simul., 8(1):3--30, Jan. 1998. Google ScholarDigital Library
- E. Pitman. Significance tests which may be applied to samples from any population. Royal Statistical Society, Supplement, 4:119--130, 1937.Google ScholarCross Ref
- J. M. Ponte and W. B. Croft. A language modeling approach to information retrieval. In Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval, SIGIR '98, pages 275--281, New York, NY, USA, 1998. ACM. Google ScholarDigital Library
- S. Robertson. Understanding inverse document frequency: On theoretical arguments IDF. Journal of Documentation, 60:503--520, 2004.Google ScholarCross Ref
- Y. Saeys, I. n. Inza, and P. Larrañaga. A review of feature selection techniques in bioinformatics. Bioinformatics, 23(19):2507--2517, Oct 2007. Google ScholarDigital Library
- M. Sanderson and J. Zobel. Information retrieval system evaluation: effort, sensitivity, and reliability. In Proceedings of the 28th annual international ACM SIGIR conference on Research and development in information retrieval, SIGIR '05, pages 162--169, New York, NY, USA, 2005. ACM. Google ScholarDigital Library
- J. Savoy. Statistical inference in retrieval effectiveness evaluation. Information Processing & Management, 33(4):495--512, 1997. Google ScholarDigital Library
- H. Scheffé. A method for judging all contrasts in the analysis of variance. Biometrika, 40(1--2):87--110, 1953.Google Scholar
- F. Scholer, A. Turpin, and M. Sanderson. Quantifying test collection quality based on the consistency of relevance judgements. In Proceedings of the 34th international ACM SIGIR conference on Research and development in Information Retrieval, SIGIR '11, pages 1063--1072, New York, NY, USA, 2011. ACM. Google ScholarDigital Library
- J. P. Shaffer. Multiple hypothesis testing. Annual Review of Psychology, 46(1):561--584, 1995.Google ScholarCross Ref
- M. D. Smucker, J. Allan, and B. Carterette. A comparison of statistical significance tests for information retrieval evaluation. In Proceedings of the sixteenth ACM conference on Conference on information and knowledge management, CIKM '07, pages 623--632, New York, NY, USA, 2007. ACM. Google ScholarDigital Library
- J. Sunklodas. Approximation of distributions of sums of weakly dependent random variables by the normal distribution. In Y. Prokhorov and V. Statulevičius, editors, Limit Theorems of Probability Theory, pages 113--165. Springer Berlin Heidelberg, 2000.Google ScholarCross Ref
- J. Tague-Sutcliffe and J. Blustein. A statistical analysis ofmboxTREC-3 data. In Overview of the Third Text REtrieval Conference (TREC-3), pages 385--398, 1994.Google Scholar
- J. Urbano, J. S. Downie, B. Mcfee, and M. Schedl. How significant is statistically significant? the case of audio music similarity and retrieval. In Proceedings of the 13th International Society for Music Information Retrieval Conference, pages 181--186, Porto, Portugal, October 8--12 2012.Google Scholar
- W. Webber, A. Moffat, and J. Zobel. Statistical power in retrieval experimentation. In Proceedings of the 17th ACM conference on Information and knowledge management, CIKM '08, pages 571--580, New York, NY, USA, 2008. ACM. Google ScholarDigital Library
- P. H. Westfall and J. F. Troendle. Multiple testing with minimal assumptions. Biometrical Journal, 50(5):745--755, 2008.Google ScholarCross Ref
- P. H. Westfall and S. S. Young. Resampling-Based Multiple Testing: Examples and Methods for p-Value Adjustment. Wiley-Interscience, 1 edition, Jan. 1993.Google Scholar
- W. J. Wilbur. Non-parametric significance tests of retrieval performance comparisons. J. Inf. Sci., 20:270--284, April 1994. Google ScholarDigital Library
- H. Xu and J. C. Hsu. Applying the generalized partitioning principle to control the generalized familywise error rate. Biometrical Journal, 49(1):52--67, 2007.Google ScholarCross Ref
- C. Zhai and J. Lafferty. A study of smoothing methods for language models applied to ad hoc information retrieval. In Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval, SIGIR '01, pages 334--342, New York, NY, USA, 2001. ACM. Google ScholarDigital Library
- J. Zhou, D. P. Foster, R. A. Stine, and L. H. Ungar. Streamwise feature selection. Journal of Machine Learning Research, 7:1861--1885, 2006. Google ScholarDigital Library
- J. Zobel, W. Webber, M. Sanderson, and A. Moffat. Principles for robust evaluation infrastructure. In Proceedings of the 2011 workshop on Data infrastructures for supporting information retrieval evaluation, DESIRE '11, pages 3--6, New York, NY, USA, 2011. ACM. Google ScholarDigital Library
Index Terms
- Deciding on an adjustment for multiplicity in IR experiments
Recommendations
A comparison of statistical significance tests for information retrieval evaluation
CIKM '07: Proceedings of the sixteenth ACM conference on Conference on information and knowledge managementInformation retrieval (IR) researchers commonly use three tests of statistical significance: the Student's paired t-test, the Wilcoxon signed rank test, and the sign test. Other researchers have previously proposed using both the bootstrap and Fisher's ...
Conducting Laboratory Experiments Properly with Statistical Tools: An Easy Hands-On Tutorial
WSDM '19: Proceedings of the Twelfth ACM International Conference on Web Search and Data MiningThis hands-on half-day tutorial consists of two sessions. Part~I covers the following topics: Preliminaries; Paired and two-sample t-tests, confidence intervals; One-way ANOVA and two-way ANOVA without replication; Familiwise error rate. Part~II covers ...
Conducting Laboratory Experiments Properly with Statistical Tools: An Easy Hands-on Tutorial
SIGIR '18: The 41st International ACM SIGIR Conference on Research & Development in Information RetrievalThis hands-on half-day tutorial consists of two 90-minute sessions. Part I covers the following topics: paired and two-sample t -tests, confidence intervals (with Excel and R); familywise error rate, multiple comparison procedures; ANOVA (with Excel and ...
Comments