Abstract:The problem of examining how well the data fit a supposed distribution is very important, and it must be confirmed prior to any data analysis, because many data analysis methods assume a specific distribution of data. For this purpose, histograms or Q-Q plots are employed for the assessment of data distribution. Additionally, a GoF TstS utilizes distance measurements between the empirical distribution function and the theoretical cumulative distribution function (cdf) to evaluate data distribution. In life-testing or reliability studies, the observed failure time of test units may not be recorded in some situations. The GoF TstSs for completely observed data can no longer be used in progressive type II censored data (PrCsD). In this paper, we suggest a GoF TstSs and new plot method for the GoF test of symmetric and asymmetric location-scale distribution (LoScD) based on PrCsD. The power of the suggested TstSs is estimated through Monte Carlo (MC) simulations, and it is compared with that of the TstSs using the order statistics (OrSt). Furthermore, we analyzed real data examples (symmetric and asymmetric data).

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in the case of Progressive Censored Data (PrCsD), how to evaluate the Goodness of Fit (GoF) of Location - Scale Distribution (LoScD). Specifically, the paper proposes a new statistic and a graphical method for evaluating the Goodness of Fit of symmetric and asymmetric Location - Scale Distributions based on Progressive Censored Data. ### Background and Problem Description In many data analysis methods, it is assumed that the data follows a specific distribution. Therefore, before any data analysis is carried out, it must be confirmed whether the data conforms to the assumed distribution. Commonly used evaluation methods include histograms, Q - Q plots, and Goodness of Fit tests based on the distance between the empirical distribution function and the theoretical cumulative distribution function (cdf). However, in life - time testing or reliability studies, in some cases, the observed failure times may not be recorded, or in order to reduce the test cost or time, some test units are removed in advance. In this case, the traditional Goodness of Fit test methods for completely observed data are no longer applicable. Therefore, the Goodness of Fit test for Progressive Censored Data has received extensive attention. ### Main Contributions of the Paper 1. **Proposing a New Statistic**: The paper proposes a new statistic based on the Lorenz Curve (LrCv) for evaluating the Goodness of Fit of Location - Scale Distribution of Progressive Censored Data. 2. **Introducing a Graphical Method**: In addition to the statistic, the paper also proposes a new graphical method for intuitively showing the fitting effect. 3. **Performance Evaluation**: Through Monte Carlo simulations, the paper evaluates the test power of the new statistic and compares it with other existing statistics. ### Formula and Symbol Explanation - **Probability Density Function (pdf) of Location - Scale Distribution**: \[ f_Y(y; \alpha, \beta)=\frac{1}{\beta} f_X\left(\frac{y - \alpha}{\beta}\right) \] where \(f_X(\cdot)\) is a known standard form, and \(\alpha\) and \(\beta\) are the location parameter and the scale parameter respectively. - **Lorenz Curve**: \[ L(p)=\frac{1}{\theta} \int_0^{F^{-1}(p)} y \, dF(y) \] where \(\theta\) is the mean of the distribution, and \(F^{-1}(p)\) is the inverse cumulative distribution function of the distribution. - **Modified Sample Lorenz Curve**: \[ mLC(p_{j:m:n})=\frac{\sum_{i = 1}^j [Y_{i:m:n}-Y_{1:m:n}]}{\sum_{i = 1}^m [Y_{i:m:n}-Y_{1:m:n}]} - p_{j:m:n}+ 1 \] - **Ratio - Modified Sample Lorenz Curve**: \[ rLC(p_{j:m:n})=\frac{mLC(p_{j:m:n})}{mLC_F(p_{j:m:n})} \] Through these formulas and methods, the paper aims to provide a more effective Goodness of Fit test tool applicable to Progressive Censored Data.

A New Test Statistic to Assess the Goodness of Fit of Location-Scale Distribution Based on Progressive Censored Data

A two-sample nonparametric test for one-sided location-scale alternative

A maximum-mean-discrepancy goodness-of-fit test for censored data

A new set of tools for goodness-of-fit validation

Generalized Residual-Based Diagnostic Testing for Duration Models with Censoring

A Self-Normalization Test for a Change-Point in the Shape Parameter of a Gamma Distributed Sequence

Kernelized Stein Discrepancy Tests of Goodness-of-fit for Time-to-Event Data

Nonparametric goodness of fit tests for Pareto type-I distribution with complete and censored data

Optimal Acceptance Sampling Testing Plan With Pivotal Quantity for Log-Location-Scale Distributions

Poisson–logarithmic half-logistic distribution with inference under a progressive-stress model based on adaptive type-II progressive hybrid censoring

Model diagnostics for censored regression via randomized survival probabilities

A goodness-of-fit test based on a recursive product of spacings

Logistic or not Logistic?

A Stein Goodness-of-fit Test for Directional Distributions

A goodness-of-fit test for testing exponentiality based on normalized dynamic survival extropy

Nonparametric Tests for the Gap Time Distributions of Serial Events Based on Censored Data

A Goodness-of-Fit Test for Statistical Models

Asymptotically Distribution-Free Goodness-of-Fit Testing: A Unifying View

Is the Gompertz family a good fit to your data?

Self-starting monitoring of the progressive type II censoring data based on goodness-of-fit test

Modified Greenwood statistic and its application for statistical testing