Multivariate coefficient of variation charts with measurement errors

Heba N Ayyoub (1) , Michael B.C. Khoo (2) , Sajal Saha (1) , Philippe Castagliola (3)

1 USM - Universiti Sains Malaysia
2 School of Mathematical Sciences
3 LS2N - équipe SLP - Systèmes Logistiques et de Production

The effects of measurement errors on the performances of multivariate charts have not been extensively investigated compared with univariate charts. More specifically, control charts for monitoring the multivariate coefficient of variation (MCV) in the existing literature are not investigated under the assumption that measurement errors are present. This research proposes two one-sided MCV charts in the presence of measurement errors, one for detecting decreasing MCV shifts and another for detecting increasing MCV shifts. The distribu-tional properties of the sample MCV in the presence of measurement errors are derived. Additionally, the for-mulae to compute the control limits and performance measures of the proposed charts are derived. Furthermore, a step-by-step procedure to implement the MCV charts with measurement errors is presented. Finally, an application of the proposed charts is illustrated using real data.

Domaines: Statistiques [stat]

Heba N Ayyoub, Michael B.C. Khoo, Sajal Saha, Philippe Castagliola. Multivariate coefficient of variation charts with measurement errors. Computers & Industrial Engineering, 2020, 147, pp.106633. ⟨10.1016/j.cie.2020.106633⟩. ⟨hal-02893090⟩