Pedagogical Issues in Hypothesis Testing (Isu-isu Pedagogi dalam Pengujian Hipotesis)

MIHIR DASH

Abstract


Hypothesis testing is a statistical technique which is used to evaluate assumptions about a population on the basis of sample data, to determine the extent to which they are tenable. Hypothesis testing is the most widely-applied statistical technique, particularly because of the emphasis on hypothesis development and testing in the scientific method. Unfortunately, students and researchers are quite prone to making mistakes and misinterpreting inferences in hypothesis testing. These mistakes and misinterpretations tend to arise from insufficient understanding of the probability and sampling theory underlying the logic of hypothesis testing. The present study attempts to identify the causes of different types of mistakes made in hypothesis testing, in order to suggest pedagogical strategies to avoid these mistakes. The data for the study was collected from a sample of postgraduate management students in Bangalore, India, using specially-designed business decision-making case lets based on hypothesis testing. The analysis focuses on the incidence of different types of mistakes that the respondents committed, particularly with respect to the type of tests, and uses multiple linear discriminant analysis to identify the factors impacting the overall inference, i.e. the correct taking of the decision and the correct drawing of the conclusion. The key finding of the study is that both the formulation and computation factors play a significant role in taking the overall inference. Further, in each panel, the critical discriminator was found to be the aspect for which the incidence of mistakes was highest. With increasing complexity of the hypothesis test, the computation factor was found to become more important. In panels A and B (tests for a single population mean and proportion, respectively), formulation aspects were found to be the most significant discriminators, and in panel C (test for equality of means), both formulation and computation aspects were significant; on the other hand, for the remaining panels (test for independence, one-way ANOVA, and two-way ANOVA), only computation aspects were significant. The study contributes to the literature by proposing some pedagogical strategies for teaching of different types of hypothesis tests based on the findings.

Keywords: Hypothesis testing; scientific method; mistakes; misinterpretations; pedagogical strategies

 

Pengujian hipotesis adalah teknik statistik yang digunakan untuk menilai andaian mengenai populasi berdasarkan data sampel, untuk mengenal pasti sejauh mana ia dapat dipertanggungjawabkan. Pengujian hipotesis adalah teknik statistik yang paling banyak digunakan, terutamanya teknik ini memberi penekanan kepada perkembangan hipotesis dan ujian dalam kaedah saintifik. Malangnya, pelajar dan penyelidik agak cenderung melakukan kesilapan dan menyalahtafsirkan kesimpulan dalam ujian hipotesis. Kesalahan dan salah tafsiran ini cenderung timbul daripada pemahaman yang tidak mencukupi tentang teori kebarangkalian dan pensampelan yang mendasari logik pengujian hipotesis. Kajian ini cuba untuk mengenal pasti punca pelbagai jenis kesilapan yang dibuat dalam pengujian hipotesis serta untuk mencadangkan strategi pedagogi bagi mengelakkan kesilapan-kesilapan ini terus berlaku. Data untuk kajian ini diambil dariapada sampel pelajar pengurusan pasca siswazah di Bangalore, India dengan menggunakan reka bentuk pembuat keputusan perniagaan yang direka khusus berdasarkan ujian hipotesis. Analisis ini memberi tumpuan kepada kejadian pelbagai jenis kesilapan yang dilakukan oleh responden, terutamanya berkenaan dengan jenis ujian, dan menggunakan analisis diskriminasi berganda untuk mengenal pasti faktor-faktor yang mempengaruhi kesimpulan keseluruhan, iaitu pengambilan keputusan yang betul dan yang betul lukisan kesimpulan. Temuan utama kajian ini adalah bahawa kedua-dua faktor perumusan dan perhitungan memainkan peranan penting dalam mengambil kesimpulan keseluruhan. Selanjutnya, dalam setiap panel, diskriminator kritikal didapati merupakan aspek kesilapan kesilapan tertinggi yang dilakukan. Dengan meningkatkannya kesukaran pengujian hipotesis, maka faktor perhitungan dijumpai menjadi lebih penting dalam pengujian hipotesis. Dalam panel A dan B (ujian bagi min dan populasi masing-masing), aspek perumusan didapati sebagai diskriminator yang paling penting. Dalam panel C (ujian untuk kesamaan sarana), aspek perumusan dan perhitungan adalah penting; Di sisi lain, untuk panel yang tinggal (ujian untuk kebebasan, ANOVA sehala, dan ANOVA dua hala), hanya aspek pengiraan yang signifikan. Kajian ini menyumbang kepada ulasna kepustakaan dengan mencadangkan beberapa strategi pedagogi untuk mengajar pelbagai jenis pengujian hipotesis berdasarkan hasil kajian.

Kata kunci: Pengujian hipotesis; kaedah sainstifik; kesilapan; salah interpretasi; strategi pedagogi


Keywords


Hypothesis testing; scientific method; mistakes; misinterpretations; pedagogical strategies

Full Text:

PDF


DOI: http://dx.doi.org/10.17576/JPEN-2018-43.01-04

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

....................................................................................................................................................................................................


Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License

© Malaysian Journal of Education | Jurnal Pendidikan Malaysia
ISSN 2180-0782 | eISSN: 2600-8823