Efficient designs for mean estimation in multilevel populations and test norming

Research output: ThesisDoctoral ThesisInternal

207 Downloads (Pure)

Abstract

A crucial step in the research process is the choice of the design of the study because a poorly designed study can have serious consequences for science (e.g. biased or unreliable results) and society (e.g. a waste of resources or bad decisions in health and education based on invalid research conclusions). This thesis deals with the design of two types of studies: surveys for mean estimation in multilevel populations (e.g. estimation of average alcohol consumption by students grouped in schools), and normative studies for estimating reference values for psychological test scores and questionnaires (e.g. to measure patients’ symptoms). Both types of studies are of practical importance: results from surveys can help policymakers, and reference values are used by clinicians or educators to assess individuals. Thus, averages and reference values must be estimated with the highest possible precision, but without wasting resources (i.e. time and money). Hence, the main objective of this thesis is to provide guidelines for planning both types of studies to achieve precise estimates using minimum resources.
Original languageEnglish
Awarding Institution
  • Maastricht University
Supervisors/Advisors
  • van Breukelen, Gerard, Supervisor
  • Candel, Math, Supervisor
  • Tan, F.E.S., Co-Supervisor
Award date20 May 2021
Place of PublicationMaastricht
Publisher
Print ISBNs9789464231809
DOIs
Publication statusPublished - 2021

Keywords

  • Optimal design
  • reference value
  • sample size calculation
  • sampling
  • statistical efficiency
  • survey

Fingerprint

Dive into the research topics of 'Efficient designs for mean estimation in multilevel populations and test norming'. Together they form a unique fingerprint.

Cite this