Reply 75 words
Hello Class, the following are my thoughts on the purpose behind hypothesis testing and the steps to follow for conducting a hypothesis test as they relate to current events.
As the textbook describes, although researchers wish to research entire populations, it is often impossible for them to do so. This impossibility is caused by a multitude of factors such as insufficient amounts of time, resources, and access. To account for this, researchers often times utilize a statistical process called hypothesis testing which allows them to use data that they gathered from a sample to make inferences about the entire population (Gravetter et al., 2021). This process involves four steps which allows researchers to gather data to reach a decision on their hypothesis.
The first step is to state the hypothesis. Within this step, the researcher will actually create two hypotheses, a null and an alternative. The null hypothesis will state that a treatment, or idea, has no effect on a population, while the alternative hypothesis states that there is a change caused by the treatment, or situation (Gravetter et al., 2021). A good example of a current event that can be used to demonstrate this is the impact on the development of young children’s social skills as a result of the shift to online learning during the pandemic. Specifically, a researcher could focus on the impact experienced by young children ages 6-8. For example, the null hypothesis could state that online learning had no effect on the development of young children’s social skills, while the alternative hypothesis could state that online learning had an impact on the development of children’s skills.
The second step in the hypothesis testing process is setting the criteria for making a decision. Within this step, researchers use data from the sample that was selected and utilize the null hypothesis for predicting where sample data will fall. Most importantly, during this step researchers will define the critical region based on the alpha level, and set the boundaries for these regions. These critical regions will be used later in the study to identify whether or not the treatment, or effect, being studied made any significant changes to the sample (Gravetter et al., 2021). Within the example above, a researcher could decide upon an alpha of .05, meaning a tail of .025 (or 2.5% resulting in a +/-1.96 z score) would be applied to the distribution on each end after z-scores are calculated. These boundaries would signify data that is significant enough to reject the null hypothesis and state that online learning actually did impact the social skills of developing children.
During step three, researchers actually begin conducting the research study and recording data. Now that a hypothesis and criteria have been established, researchers are able to remain unbiased while they collect their raw data and take measurements. During this stage z-scores are also calculated which will help researchers identify differences between the data and hypothesis and plot the new average to determine if the treatment, or studied effect, made an impact (Gravetter et al., 2021). Within the example, researchers can utilize a developmental test focusing on social skills to measure children, aged 6-8, who went through online school during the pandemic. This data can then be utilized to determine a mean, standard deviation, and associated z-scores.
Lastly, during step four, researchers will make a decision on their hypothesis. This decision is based upon the information, mainly the z-scores, that were obtained and calculated during step three. The z-scores will be plotted on a distribution scale to determine if the scores taken fell within the critical regions, which were determined during step two. If scores fall within the critical region, a researcher would likely decide to reject the null hypothesis and accept the alternative. If the scores did not fall within the critical region, the null hypothesis would not be rejected (Gravetter et al., 2021). For the example listed above, if during step three, the scores gathered from the children resulted in an average z-score of 1.43, then it would fail to fall within the critical region (1.96). As a result, the researcher would have to decide to reject the null hypothesis and support the idea that online learning did not impact the development of social skills for young children aged 6-8.