Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

Sorted by: relevance · author · university · dateNew search

You searched for +publisher:"Kansas State University" +contributor:("Paul I. Nelson"). Showing records 1 – 12 of 12 total matches.

Search Limiters

Last 2 Years | English Only

No search limiters apply to these results.

▼ Search Limiters

1. Dey, Rajarshi. Inference for the K-sample problem based on precedence probabilities.

Degree: PhD, Department of Statistics, 2012, Kansas State University

 Rank based inference using independent random samples to compare K>1 continuous distributions, called the K-sample problem, based on precedence probabilities is developed and explored. There… (more)

Subjects/Keywords: Precedence probabilities; Nonlinear rank-based statistic; U-statistic; K-sample problem; Index measures of distance among distributions; Hypervolume under ROC manifold; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Dey, R. (2012). Inference for the K-sample problem based on precedence probabilities. (Doctoral Dissertation). Kansas State University. Retrieved from http://hdl.handle.net/2097/12000

Chicago Manual of Style (16th Edition):

Dey, Rajarshi. “Inference for the K-sample problem based on precedence probabilities.” 2012. Doctoral Dissertation, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/12000.

MLA Handbook (7th Edition):

Dey, Rajarshi. “Inference for the K-sample problem based on precedence probabilities.” 2012. Web. 24 Jan 2020.

Vancouver:

Dey R. Inference for the K-sample problem based on precedence probabilities. [Internet] [Doctoral dissertation]. Kansas State University; 2012. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/12000.

Council of Science Editors:

Dey R. Inference for the K-sample problem based on precedence probabilities. [Doctoral Dissertation]. Kansas State University; 2012. Available from: http://hdl.handle.net/2097/12000

2. Hua, Jianjun. Confidence intervals for population size based on a capture-recapture design.

Degree: MS, Department of Statistics, 2011, Kansas State University

 Capture-Recaputre (CR) experiments stemmed from the study of wildlife and are widely used in areas such as ecology, epidemiology, evaluation of census undercounts, and software… (more)

Subjects/Keywords: Capture-recapture; Confidence interval; Population size; Asymptotic normality estimation; Inverting a test; Coverage rate; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Hua, J. (2011). Confidence intervals for population size based on a capture-recapture design. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/13103

Chicago Manual of Style (16th Edition):

Hua, Jianjun. “Confidence intervals for population size based on a capture-recapture design.” 2011. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/13103.

MLA Handbook (7th Edition):

Hua, Jianjun. “Confidence intervals for population size based on a capture-recapture design.” 2011. Web. 24 Jan 2020.

Vancouver:

Hua J. Confidence intervals for population size based on a capture-recapture design. [Internet] [Masters thesis]. Kansas State University; 2011. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/13103.

Council of Science Editors:

Hua J. Confidence intervals for population size based on a capture-recapture design. [Masters Thesis]. Kansas State University; 2011. Available from: http://hdl.handle.net/2097/13103


Kansas State University

3. Ling, Yan. Inference for the intrinsic separation among distributions which may differ in location and scale.

Degree: PhD, Department of Statistics, 2009, Kansas State University

 The null hypothesis of equal distributions, H0 : F1[equals]F2[equals]...[equals]FK , is commonly used to compare two or more treatments based on data consisting of independent… (more)

Subjects/Keywords: intrinsic separation parameter (ISP); normal distribution; nuisance parameter; p-value; average p-value; parametric bootstrap test; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Ling, Y. (2009). Inference for the intrinsic separation among distributions which may differ in location and scale. (Doctoral Dissertation). Kansas State University. Retrieved from http://hdl.handle.net/2097/1845

Chicago Manual of Style (16th Edition):

Ling, Yan. “Inference for the intrinsic separation among distributions which may differ in location and scale.” 2009. Doctoral Dissertation, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/1845.

MLA Handbook (7th Edition):

Ling, Yan. “Inference for the intrinsic separation among distributions which may differ in location and scale.” 2009. Web. 24 Jan 2020.

Vancouver:

Ling Y. Inference for the intrinsic separation among distributions which may differ in location and scale. [Internet] [Doctoral dissertation]. Kansas State University; 2009. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/1845.

Council of Science Editors:

Ling Y. Inference for the intrinsic separation among distributions which may differ in location and scale. [Doctoral Dissertation]. Kansas State University; 2009. Available from: http://hdl.handle.net/2097/1845


Kansas State University

4. Gayle, Suelen S. A simulation study of the size and power of Cochran’s Q versus the standard Chi-square test for testing the equality of correlated proportions.

Degree: MS, Department of Statistics, 2010, Kansas State University

 The standard Chi-square test for the equality of proportions of positive responses to c specified binary questions is valid when the observed responses arise from… (more)

Subjects/Keywords: Cochran's Q; Chi-Square; Binary Correlated Proportions; Power; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Gayle, S. S. (2010). A simulation study of the size and power of Cochran’s Q versus the standard Chi-square test for testing the equality of correlated proportions. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/3881

Chicago Manual of Style (16th Edition):

Gayle, Suelen S. “A simulation study of the size and power of Cochran’s Q versus the standard Chi-square test for testing the equality of correlated proportions.” 2010. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/3881.

MLA Handbook (7th Edition):

Gayle, Suelen S. “A simulation study of the size and power of Cochran’s Q versus the standard Chi-square test for testing the equality of correlated proportions.” 2010. Web. 24 Jan 2020.

Vancouver:

Gayle SS. A simulation study of the size and power of Cochran’s Q versus the standard Chi-square test for testing the equality of correlated proportions. [Internet] [Masters thesis]. Kansas State University; 2010. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/3881.

Council of Science Editors:

Gayle SS. A simulation study of the size and power of Cochran’s Q versus the standard Chi-square test for testing the equality of correlated proportions. [Masters Thesis]. Kansas State University; 2010. Available from: http://hdl.handle.net/2097/3881

5. Serasinghe, Shyamalee Kumary. A simulation comparison of parametric and nonparametric estimators of quantiles from right censored data.

Degree: MS, Department of Statistics, 2010, Kansas State University

 Quantiles are useful in describing distributions of component lifetimes. Data, consisting of the lifetimes of sample units, used to estimate quantiles are often censored. Right… (more)

Subjects/Keywords: Estimating Quantiles; Right Censored Data; Generating Right Censored Data; Kaplan-Meier Estimator; Parametric Estimators of Quantiles; Nonparametric Estimators of Quantiles; Statistics (0463)

Page 1 Page 2 Page 3 Page 4 Page 5

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Serasinghe, S. K. (2010). A simulation comparison of parametric and nonparametric estimators of quantiles from right censored data. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/4318

Chicago Manual of Style (16th Edition):

Serasinghe, Shyamalee Kumary. “A simulation comparison of parametric and nonparametric estimators of quantiles from right censored data.” 2010. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/4318.

MLA Handbook (7th Edition):

Serasinghe, Shyamalee Kumary. “A simulation comparison of parametric and nonparametric estimators of quantiles from right censored data.” 2010. Web. 24 Jan 2020.

Vancouver:

Serasinghe SK. A simulation comparison of parametric and nonparametric estimators of quantiles from right censored data. [Internet] [Masters thesis]. Kansas State University; 2010. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/4318.

Council of Science Editors:

Serasinghe SK. A simulation comparison of parametric and nonparametric estimators of quantiles from right censored data. [Masters Thesis]. Kansas State University; 2010. Available from: http://hdl.handle.net/2097/4318


Kansas State University

6. Paranagama, Thilanka Dilruwani. A simulation study of the robustness of the least median of squares estimator of slope in a regression through the origin model.

Degree: MS, Department of Statistics, 2010, Kansas State University

 The principle of least squares applied to regression models estimates parameters by minimizing the mean of squared residuals. Least squares estimators are optimal under normality… (more)

Subjects/Keywords: Least median of squares estimates; Regression; Estimates; Median; Regression through the origin; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Paranagama, T. D. (2010). A simulation study of the robustness of the least median of squares estimator of slope in a regression through the origin model. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/7045

Chicago Manual of Style (16th Edition):

Paranagama, Thilanka Dilruwani. “A simulation study of the robustness of the least median of squares estimator of slope in a regression through the origin model.” 2010. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/7045.

MLA Handbook (7th Edition):

Paranagama, Thilanka Dilruwani. “A simulation study of the robustness of the least median of squares estimator of slope in a regression through the origin model.” 2010. Web. 24 Jan 2020.

Vancouver:

Paranagama TD. A simulation study of the robustness of the least median of squares estimator of slope in a regression through the origin model. [Internet] [Masters thesis]. Kansas State University; 2010. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/7045.

Council of Science Editors:

Paranagama TD. A simulation study of the robustness of the least median of squares estimator of slope in a regression through the origin model. [Masters Thesis]. Kansas State University; 2010. Available from: http://hdl.handle.net/2097/7045

7. Devamitta Perera, Muditha Virangika. Robustness of normal theory inference when random effects are not normally distributed.

Degree: MS, Department of Statistics, 2011, Kansas State University

 The variance of a response in a one-way random effects model can be expressed as the sum of the variability among and within treatment levels.… (more)

Subjects/Keywords: Random Effects Models; Non-normal random effects; Variance Components; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Devamitta Perera, M. V. (2011). Robustness of normal theory inference when random effects are not normally distributed. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/8786

Chicago Manual of Style (16th Edition):

Devamitta Perera, Muditha Virangika. “Robustness of normal theory inference when random effects are not normally distributed.” 2011. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/8786.

MLA Handbook (7th Edition):

Devamitta Perera, Muditha Virangika. “Robustness of normal theory inference when random effects are not normally distributed.” 2011. Web. 24 Jan 2020.

Vancouver:

Devamitta Perera MV. Robustness of normal theory inference when random effects are not normally distributed. [Internet] [Masters thesis]. Kansas State University; 2011. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/8786.

Council of Science Editors:

Devamitta Perera MV. Robustness of normal theory inference when random effects are not normally distributed. [Masters Thesis]. Kansas State University; 2011. Available from: http://hdl.handle.net/2097/8786


Kansas State University

8. Wu, Yun. A simulation study of the robustness of Hotelling’s T2 test for the mean of a multivariate distribution when sampling from a multivariate skew-normal distribution.

Degree: MS, Department of Statistics, 2009, Kansas State University

 Hotelling’s T2 test is the standard tool for inference about the mean of a multivariate normal population. However, this test may perform poorly when used… (more)

Subjects/Keywords: Simulation study; Hotelling's T2 test; Skew normal; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Wu, Y. (2009). A simulation study of the robustness of Hotelling’s T2 test for the mean of a multivariate distribution when sampling from a multivariate skew-normal distribution. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/1939

Chicago Manual of Style (16th Edition):

Wu, Yun. “A simulation study of the robustness of Hotelling’s T2 test for the mean of a multivariate distribution when sampling from a multivariate skew-normal distribution.” 2009. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/1939.

MLA Handbook (7th Edition):

Wu, Yun. “A simulation study of the robustness of Hotelling’s T2 test for the mean of a multivariate distribution when sampling from a multivariate skew-normal distribution.” 2009. Web. 24 Jan 2020.

Vancouver:

Wu Y. A simulation study of the robustness of Hotelling’s T2 test for the mean of a multivariate distribution when sampling from a multivariate skew-normal distribution. [Internet] [Masters thesis]. Kansas State University; 2009. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/1939.

Council of Science Editors:

Wu Y. A simulation study of the robustness of Hotelling’s T2 test for the mean of a multivariate distribution when sampling from a multivariate skew-normal distribution. [Masters Thesis]. Kansas State University; 2009. Available from: http://hdl.handle.net/2097/1939

9. Makarova, Natalya. A simulation study of the robustness of prediction intervals for an independent observation obtained from a random sample from an assumed location-scale family of distributions.

Degree: MS, Department of Statistics, 2012, Kansas State University

 Suppose that based on data consisting of independent repetitions of an experiment a researcher wants to predict the outcome of the next independent outcome of… (more)

Subjects/Keywords: Prediction intervals; Robustness; Simulation study; Location-scale family of distributions; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Makarova, N. (2012). A simulation study of the robustness of prediction intervals for an independent observation obtained from a random sample from an assumed location-scale family of distributions. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/14749

Chicago Manual of Style (16th Edition):

Makarova, Natalya. “A simulation study of the robustness of prediction intervals for an independent observation obtained from a random sample from an assumed location-scale family of distributions.” 2012. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/14749.

MLA Handbook (7th Edition):

Makarova, Natalya. “A simulation study of the robustness of prediction intervals for an independent observation obtained from a random sample from an assumed location-scale family of distributions.” 2012. Web. 24 Jan 2020.

Vancouver:

Makarova N. A simulation study of the robustness of prediction intervals for an independent observation obtained from a random sample from an assumed location-scale family of distributions. [Internet] [Masters thesis]. Kansas State University; 2012. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/14749.

Council of Science Editors:

Makarova N. A simulation study of the robustness of prediction intervals for an independent observation obtained from a random sample from an assumed location-scale family of distributions. [Masters Thesis]. Kansas State University; 2012. Available from: http://hdl.handle.net/2097/14749

10. Jiao, Yuanfang. A simulation comparison of two methods for controlling the experiment-wise Type I error rate of correlated tests for contrasts in one-way completely randomized designs.

Degree: MS, Department of Statistics, 2014, Kansas State University

 A Bonferroni and an ordered P-value solution to the problem of controlling the experiment-wise Type I error rate are studied and compared in terms of… (more)

Subjects/Keywords: Experiment-wise type I error; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Jiao, Y. (2014). A simulation comparison of two methods for controlling the experiment-wise Type I error rate of correlated tests for contrasts in one-way completely randomized designs. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/17619

Chicago Manual of Style (16th Edition):

Jiao, Yuanfang. “A simulation comparison of two methods for controlling the experiment-wise Type I error rate of correlated tests for contrasts in one-way completely randomized designs.” 2014. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/17619.

MLA Handbook (7th Edition):

Jiao, Yuanfang. “A simulation comparison of two methods for controlling the experiment-wise Type I error rate of correlated tests for contrasts in one-way completely randomized designs.” 2014. Web. 24 Jan 2020.

Vancouver:

Jiao Y. A simulation comparison of two methods for controlling the experiment-wise Type I error rate of correlated tests for contrasts in one-way completely randomized designs. [Internet] [Masters thesis]. Kansas State University; 2014. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/17619.

Council of Science Editors:

Jiao Y. A simulation comparison of two methods for controlling the experiment-wise Type I error rate of correlated tests for contrasts in one-way completely randomized designs. [Masters Thesis]. Kansas State University; 2014. Available from: http://hdl.handle.net/2097/17619


Kansas State University

11. Yankovskyy, Yevhen. Application of a Gibbs Sampler to estimating parameters of a hierarchical normal model with a time trend and testing for existence of the global warming.

Degree: MS, Department of Statistics, 2008, Kansas State University

 This research is devoted to studying statistical inference implemented using the Gibbs Sampler for a hierarchical Bayesian linear model with first order autoregressive structure. This… (more)

Subjects/Keywords: hierarchical; model; Gibbs; Sampler; global; warming; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Yankovskyy, Y. (2008). Application of a Gibbs Sampler to estimating parameters of a hierarchical normal model with a time trend and testing for existence of the global warming. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/1010

Chicago Manual of Style (16th Edition):

Yankovskyy, Yevhen. “Application of a Gibbs Sampler to estimating parameters of a hierarchical normal model with a time trend and testing for existence of the global warming.” 2008. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/1010.

MLA Handbook (7th Edition):

Yankovskyy, Yevhen. “Application of a Gibbs Sampler to estimating parameters of a hierarchical normal model with a time trend and testing for existence of the global warming.” 2008. Web. 24 Jan 2020.

Vancouver:

Yankovskyy Y. Application of a Gibbs Sampler to estimating parameters of a hierarchical normal model with a time trend and testing for existence of the global warming. [Internet] [Masters thesis]. Kansas State University; 2008. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/1010.

Council of Science Editors:

Yankovskyy Y. Application of a Gibbs Sampler to estimating parameters of a hierarchical normal model with a time trend and testing for existence of the global warming. [Masters Thesis]. Kansas State University; 2008. Available from: http://hdl.handle.net/2097/1010


Kansas State University

12. Yang, Celeste. A study of the calibration-inverse prediction problem in a mixed model setting.

Degree: MS, Department of Statistics, 2008, Kansas State University

 The Calibration-Inverse Prediction Problem was investigated in a mixed model setting. Two methods were used to construct inverse prediction intervals. Method 1 ignores the random… (more)

Subjects/Keywords: Calibration; Inverse Prediction; Mixed Model; RTLA; Statistics (0463)

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Yang, C. (2008). A study of the calibration-inverse prediction problem in a mixed model setting. (Masters Thesis). Kansas State University. Retrieved from http://hdl.handle.net/2097/1079

Chicago Manual of Style (16th Edition):

Yang, Celeste. “A study of the calibration-inverse prediction problem in a mixed model setting.” 2008. Masters Thesis, Kansas State University. Accessed January 24, 2020. http://hdl.handle.net/2097/1079.

MLA Handbook (7th Edition):

Yang, Celeste. “A study of the calibration-inverse prediction problem in a mixed model setting.” 2008. Web. 24 Jan 2020.

Vancouver:

Yang C. A study of the calibration-inverse prediction problem in a mixed model setting. [Internet] [Masters thesis]. Kansas State University; 2008. [cited 2020 Jan 24]. Available from: http://hdl.handle.net/2097/1079.

Council of Science Editors:

Yang C. A study of the calibration-inverse prediction problem in a mixed model setting. [Masters Thesis]. Kansas State University; 2008. Available from: http://hdl.handle.net/2097/1079

.