Hypothesis is a term that is often used in many ways (for example, sometimes it means that you have a feeling that something is true), but it has a precise meaning in statistics. Hypotheses are the likely answers to our testable research questions. Hypotheses are derived from prior knowledge and questions that arise from previously conducted research – they represent what people still need to learn about a particular area (Wagner III & Gillespie, 2019).
For example, you know that water is required to sustain life, and you recently discovered snow on Mars. Now, astronomers are exploring the research question, “Can Mars sustain human life?” Their hypothesis could be, “If there is water on Mars, then humans will be able to survive.” Notice that this is a testable statement! Through observation or experiments, you can determine whether this hypothesis is true or false. Notice, too, that the hypothesis indicates the relation between the independent variable (available water) and the dependent variable (sustaining human life). This week, you will learn to write strong hypotheses for statistical analyses.
You will also learn about statistical errors. Since it is not possible to be completely confident in the results of a study, researchers accept – but try to reduce – the probability of error in their conclusions. When testing a hypothesis, your first step will be to name the null hypothesis, which is typically the default state. This is done within this nation’s legal system all the time – the default state, or the null hypothesis, is that any accused person is not guilty. Then, people test that hypothesis through the process of an investigation and a trial. The result of this test may be that people reject the null hypothesis – or replace the default state (‘this person is not guilty’) with something else (‘this person is guilty’). If the result of this test corresponds with reality, then individuals made the correct decision to reject the null hypothesis. However, if the result of the test does not correspond with reality, then an error has occurred. There are two types of error that you will explore this week: Type I error, also referred to as “false positive”, when you reject the null hypothesis, but it is true (in this example, you claim that a person is guilty, but in reality, the person is not) and Type II error, also known as “false negative”, when you keep the null hypothesis though it is false (you find the accused person innocent, but in reality, that person did commit the crime). You can see that each type of error has consequences; you will be introduced to guidelines for statistical testing that can help you limit the errors in your analyses.
This week, you will also explore inferential statistics like T test, which will help in hypothesis testing and arriving at conclusions about the hypothesis. Inferential statistics use a random sample from a population to describe a population, thus being able to make inferences about the population as a whole. A T test is used to compare two groups of means to determine if there is a statistically significant difference between the groups. The basic premise behind the T test statistic is to compare distributions based on the difference between means. In the simplest terms, the logic is to calculate the difference between means (Wagner III & Gillespie, 2019).
You will learn that there are different types of T tests like One sample T test, Paired Sample T test, and Independent sample T test that can be used depending on the research design. Importance is placed on calculating these tests by using statistical software, as well as analyzing and interpreting their results. In addition to manipulating computations, the conceptual basis of each test will be explored.
References
Wagner III, W., & Gillespie, B. (2019). Using and interpreting statistics in the social, behavioral, and health sciences. Thousand Oaks, CA: SAGE.
Weekly Resources and Assignments
Review the resources from the Course Resources link, located in the top navigation bar, to prepare for this week’s assignments. The resources may include textbook reading assignments, journal articles, websites, links to tools or software, videos, handouts, rubrics, etc.
0 %0 of 1 topics completeShow data table for This chart displays the number of completed topics versus the total number of topics within module Week 6..
List of Topics and Sub-Modules for Week 6