Article Preview
Top1. Introduction
Teaching inferential statistics is one of the most tedious tasks for statistics educators (Park, 2019). Often, students cannot create logical connections between inferential statistics and other subjects, such as descriptive statistics and probability (Park, 2019). The root of such difficulty may be that inferential statistics is anchored on complex constructs such as hypothetical reasoning, data analytic methods, and probabilistic thinking (Park, 2019). To address this issue, several improvements in the school curriculum were proposed in the current literature (e.g., National Council of Teachers of Mathematics, 2000; Batanero et al., 2011; Franklin et al. 2007; Park, 2019). Despite the initiatives to improve the school curriculum, inferential statistics persist in being very challenging to most students as well as to adults (Park, 2019).
Hypothesis testing (HT) (also called significance test) is the primary tool in inferential statistics; however, among formal statistical inference topics, it is highly challenging to learn and apply (Sotos et al., 2007; Makar, 2016). This technique’s goal is to define the evidence in a sample versus a previously defined (null) hypothesis, minimizing certain risks. Getting students to make sense of hypothesis tests is a challenging goal for statistics instructors (Sotos et al., 2007). Park (2019) broke down the complexity of learning HT into procedural components using the conventional teaching method. These predetermined procedures enabled students to perform inferential statistics without reinventing the wheel quickly. However, the conventional teaching method limits the students’ critical thinking (Park, 2019). Although students were able to perform the procedures surrounding HT, they may lack a strong understanding of the concepts and their use, which makes the learning not holistic. Holistic learning of HT requires students not only to know the procedure for performing the test, but also to understand and be able to relate many abstract concepts such as the concept of a sampling distribution, the significance level, null and alternative hypotheses, the 
value, and so on (Kirk, 2001; Sotos et al., 2007; Makar, 2016). In this regard, conventional teaching methods can lack flexibility, do not ensure teaching consistency, nor accommodate the different learning needs of students (Jeffries, 2001; Jeffries et al., 2002; Bloomfield et al., 2010).
A promising method that can be utilized to teach HT instead of conventional teaching methods effectively is teaching via learning modality, which was first posited in Dunn and Dunn’s (1978) learning style model. Learning styles, learning preferences, learning modality, and cognitive styles all refer to individual differences in preference for a specific approach to learning. The idea that each student has a particular learning modality is one of the most resilient topics in all of education (Lodge et al., 2015) and that students will learn more efficiently when a teaching method matches their preferred style (Kozhevnikov et al., 2014).