The Differences Among Methods for Computation of Quartiles Do Matter
Darja Rupnik Poklukar and Janez Zerovnik
University of Ljubljana, Faculty of Mechanical Engineering, Askerceva 6, SI-1000 Ljubljana, Slovenia
Abstract: The choice of the method for computation of quartiles may have a direct impact on practical decisions. As the differences among the methods tend to vanish with growing sample size, a common belief is that the methods are practically equivalent. To the contrary, we show on several experiments with various sample sizes that in some examples, the differences may be very likely. For a discrete distribution, exact estimates of differences between expected values of sample quartiles, given by two different methods, are derived. This implies that it is crucial that in any application, the choice of the method for computation of quartiles (or, percentiles) is explicitly given.
Keywords: quantiles, percentiles, sample quartiles.
Pages: 379 – 388 | Full PDF Paper
Education on Statistics for Public Health Nurses at a Statistics Seminar
Toshiharu ETO1, Keiko AOISHI2, Noriko MATSUMOTO1, Yoshio HIDAKA3, Yoshinori FUJII4 and Hiroyuki NAKAO1
1. Miyazaki prefectural nursing university, Faculty of nursing, 3-5-1 Manabino, Miyazaki, 880-0929, JAPAN
2. Nagoya university graduate school of Medicine, 1-1-20, Daikouminami, Higashi, Nagoya, Aichi, 461-8673, JAPAN
3. Miyazaki prefectural office Division of Health & Welfare, 2-10-1 Tachibanadori Higashi, 880-8501, JAPAN
4. University of Miyazaki, Faculty of education, 1-1 Gakuenkibanadai-nishi, Miyazaki, 889-2192, JAPAN
Abstract: It is quite important to effectively develop community healthcare services and health promotion activities in an aged society. Therefore, statistics for understanding local characteristics is essential for evidence-based healthcare services and healthy-policy decision in a community. We started a “HIMUKA health research seminar” project in April 2015. The purposes of the seminar are education and training of the study driving force, enhancement of knowledge and capability of public health nurses to utilize statistics for healthy-policy decision. The total number of seminar participants was 545 in 2 years. The major motivations for participating in the seminar were their interest in “Analytical methods using Excel”, “Visualization of data by creating tables and figures”, “How to work out a questionnaire”, and “How to reflect analyzed data to a healthy policy”.
Keywords: Education on statics, Public health nurses,Compound seminar
Pages: 389 – 396 | Full PDF Paper
Igniting the Statistical Spark in the Social Sciences
Rhys Christopher Jones
Department of Statistics, University of Auckland, Auckland 1010.
Abstract: The Cardiff University Q-Step centre has become a national leader in developing and promoting context rich statistical courses (aimed at ages 16-18), across England and Wales for the past 8 years. These courses involve a new subject area called Social Analytics (SA) (the scientific investigation of social processes using statistical techniques and analysis). This article will focus on the development of the courses interdisciplinary curriculum, as well as a description of the pedagogical methods adopted, underpinned by a constructivist approach to learning. There will also be a brief overview of a quasi-experimental approach utilised to evaluate the impact of the course on year 12 and 13 attitudes and abilities in mathematics, statistics and critical thinking, compared to two control groups.
Pages: 397 – 403 | Full PDF Paper