Statistical analysis of ordinal user opinion scores

Mu Mu; Andreas Mauthe; Gareth Tyson; Eduardo Cerqueira

doi:10.1109/CCNC.2012.6181112

Statistical analysis of ordinal user opinion scores

Mu Mu^*, Andreas Mauthe, Gareth Tyson, Eduardo Cerqueira

^*Corresponding author for this work

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

10 Citations (Scopus)

Abstract

Data-sets derived from subjective experiments are often exploited to construct objective quality models using parametric statistics such as MOS and multiple regression. In this paper, using data type and normality tests, we verify that non-normally distributed user opinion scores in nominal or ordinal responses should not be analysed using parametric statistics. The paper introduces a number of non-parametric statistics for valid model building and parameter estimation based on user opinion scores. A set of modelling results are also presented to demonstrate the effectiveness of non-parametric statistics.

Original language	English
Title of host publication	2012 IEEE Consumer Communications and Networking Conference, CCNC'2012
Pages	331-336
Number of pages	6
DOIs	https://doi.org/10.1109/CCNC.2012.6181112
Publication status	Published - 2012
Externally published	Yes
Event	2012 IEEE Consumer Communications and Networking Conference, CCNC'2012 - Las Vegas, NV, United States Duration: 14 Jan 2012 → 17 Jan 2012

Publication series

Name	2012 IEEE Consumer Communications and Networking Conference, CCNC'2012

Conference

Conference	2012 IEEE Consumer Communications and Networking Conference, CCNC'2012
Country/Territory	United States
City	Las Vegas, NV
Period	14/01/12 → 17/01/12

Access to Document

10.1109/CCNC.2012.6181112

Cite this

@inproceedings{9a30e92244624a0f82f121e2b99b4f11,

title = "Statistical analysis of ordinal user opinion scores",

abstract = "Data-sets derived from subjective experiments are often exploited to construct objective quality models using parametric statistics such as MOS and multiple regression. In this paper, using data type and normality tests, we verify that non-normally distributed user opinion scores in nominal or ordinal responses should not be analysed using parametric statistics. The paper introduces a number of non-parametric statistics for valid model building and parameter estimation based on user opinion scores. A set of modelling results are also presented to demonstrate the effectiveness of non-parametric statistics.",

author = "Mu Mu and Andreas Mauthe and Gareth Tyson and Eduardo Cerqueira",

year = "2012",

doi = "10.1109/CCNC.2012.6181112",

language = "English",

isbn = "9781457720710",

series = "2012 IEEE Consumer Communications and Networking Conference, CCNC'2012",

pages = "331--336",

booktitle = "2012 IEEE Consumer Communications and Networking Conference, CCNC'2012",

note = "2012 IEEE Consumer Communications and Networking Conference, CCNC'2012 ; Conference date: 14-01-2012 Through 17-01-2012",

}

Mu, M, Mauthe, A, Tyson, G & Cerqueira, E 2012, Statistical analysis of ordinal user opinion scores. in 2012 IEEE Consumer Communications and Networking Conference, CCNC'2012., 6181112, 2012 IEEE Consumer Communications and Networking Conference, CCNC'2012, pp. 331-336, 2012 IEEE Consumer Communications and Networking Conference, CCNC'2012, Las Vegas, NV, United States, 14/01/12. https://doi.org/10.1109/CCNC.2012.6181112

Statistical analysis of ordinal user opinion scores. / Mu, Mu; Mauthe, Andreas; Tyson, Gareth et al.
2012 IEEE Consumer Communications and Networking Conference, CCNC'2012. 2012. p. 331-336 6181112 (2012 IEEE Consumer Communications and Networking Conference, CCNC'2012).

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

TY - GEN

T1 - Statistical analysis of ordinal user opinion scores

AU - Mu, Mu

AU - Mauthe, Andreas

AU - Tyson, Gareth

AU - Cerqueira, Eduardo

PY - 2012

Y1 - 2012

N2 - Data-sets derived from subjective experiments are often exploited to construct objective quality models using parametric statistics such as MOS and multiple regression. In this paper, using data type and normality tests, we verify that non-normally distributed user opinion scores in nominal or ordinal responses should not be analysed using parametric statistics. The paper introduces a number of non-parametric statistics for valid model building and parameter estimation based on user opinion scores. A set of modelling results are also presented to demonstrate the effectiveness of non-parametric statistics.

AB - Data-sets derived from subjective experiments are often exploited to construct objective quality models using parametric statistics such as MOS and multiple regression. In this paper, using data type and normality tests, we verify that non-normally distributed user opinion scores in nominal or ordinal responses should not be analysed using parametric statistics. The paper introduces a number of non-parametric statistics for valid model building and parameter estimation based on user opinion scores. A set of modelling results are also presented to demonstrate the effectiveness of non-parametric statistics.

UR - https://www.scopus.com/pages/publications/84860671374

U2 - 10.1109/CCNC.2012.6181112

DO - 10.1109/CCNC.2012.6181112

M3 - Conference Paper published in a book

AN - SCOPUS:84860671374

SN - 9781457720710

T3 - 2012 IEEE Consumer Communications and Networking Conference, CCNC'2012

SP - 331

EP - 336

BT - 2012 IEEE Consumer Communications and Networking Conference, CCNC'2012

T2 - 2012 IEEE Consumer Communications and Networking Conference, CCNC'2012

Y2 - 14 January 2012 through 17 January 2012

ER -

Statistical analysis of ordinal user opinion scores

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this