Comparing experimental conditions using modern statistics
- PDF / 2,055,581 Bytes
- 22 Pages / 595.224 x 790.955 pts Page_size
- 40 Downloads / 193 Views
Comparing experimental conditions using modern statistics Jean-Bernard Martens1
© The Author(s) 2020
Abstract While the applied psychology community relies on statistics to assist drawing conclusions from quantitative data, the methods being used mostly today do not reflect several of the advances in statistics that have been realized over the past decades. We show in this paper how a number of issues with how statistical analyses are presently executed and reported in the literature can be addressed by applying more modern methods. Unfortunately, such new methods are not always supported by widely available statistical packages, such as SPSS, which is why we also introduce a new software platform, called ILLMO (for Interactive Log-Likelihood MOdeling), which offers an intuitive interface to such modern statistical methods. In order to limit the complexity of the material being covered in this paper, we focus the discussion on a fairly simple, but nevertheless very frequent and important statistical task, i.e., comparing two experimental conditions. Keywords Interactive statistics · Exploratory statistics · Hypothesis testing · Effect size · t test · Likert scales · Confidence intervals · Wilks’ theorem · Empirical likelihood
Introduction The applied psychology community, including my own field of research in human–computer interaction (HCI), relies heavily on empirical research to validate the claims that they make. This empirical research often involves a mix of qualitative and quantitative methods, and statistics is used to analyse the data generated by the quantitative methods. One of the most frequently occurring tasks is to compare two experimental conditions, where the two conditions for instance correspond to a proposed intervention being absent or present. The quantitative measure being used to compare such conditions can either be objective, such as measuring performance time or number of mistakes made, or subjective, such as letting participants express their assessment on one or more attributes on a (7-point) Likert scale. Such measurements can either be performed by the same participants, who get to experience both conditions (i.e., a within-subject experiment), or by two separate groups of participants, each group experiencing one of the conditions (i.e., an across-subject experiment). We will use this relatively simple experimental setup to identify some Jean-Bernard Martens
[email protected] 1
Eindhoven University of Technology, Eindhoven, The Netherlands
important issues with how statistical analyses are currently performed and reported, and to propose concrete ways of making improvements by incorporating more modern statistical methods. In this paper, we will use example data from the popular book by Andy Field (Field, 2013), as this source also provides example formulations for how to report the statistical analyses that are traditionally performed. We will see that these traditional methods focus on establishing statistical significance, instead of on the more relevant aspect of est
Data Loading...