Null Hypothesis Experiment Resources - Assignment Example

Running head: QUESTIONS Questions Affiliation. Questions Part I Distinguish between the null hypothesis and the research hypothesis. When does the researcher decide to reject the null hypothesis? The null hypothesis is something that the researcher is actually trying to disprove or reject. A researcher rejects the null hypothesis when he or she can compile enough data to show that it is at least 95% untrue (or untrue 95% of the time). (Null Hypothesis experiment resources, n.d.) 2. What is meant by statistical significance? Statistical significance means that the statistical results are reliable. However, it does not necessarily mean that the finding has any decision-making utility. (Statistical Significance, 1997) 3. What factors are most important in determining whether obtained results will be satisfied? Obtained results are the results that the test demonstrates overall. The factors that are most important when determining whether these results will be satisfied are the hypothesis, the research itself, and the statistical significance of the research. 4. Distinguish between a Type I and a Type II error. Why is your significance level the probability of making a Type I error? Type I and Type II errors are both types of errors that can be made in significance testing. A Type I error occurs when a null hypothesis is rejected, and it should not have been rejected. A Type II error occurs when a false null hypothesis is not rejected. The probability of a Type I error is designated by the Greek letter alpha (a) and is called the Type I error rate; the probability of a Type II error (the Type II error rate) is designated by the Greek letter beta (ß) . The false positive rate can be defined, according to Type I and Type II Errors (n.d.) as: the proportion of negative instances that were erroneously reported as being positive. It is equal to 1 minus the specificity of the test. This is equivalent to saying the false positive rate is equal to the significance level. This means the false positive rate = number of false positives/total number of negative instances. 5. What factors are involved in choosing a significance level? “Statistical Significance” states (1997): Decide on the critical alpha level you will use (i.e., the error rate you are willing to accept). Conduct the research. Calculate the statistic. Compare the statistic to a critical value obtained from a table. If your statistic is higher than the critical value from the table: Your finding is significant. You reject the null hypothesis. The probability is small that the difference or relationship happened by chance, and p is l less than the critical alpha level (p < alpha ). If your statistic is lower than the critical value from the table: Your finding is not significant. You fail to reject the null hypothesis. The probability is high that the difference or relationship happened by chance, and p is greater than the critical alpha level (p > alpha ) (para 6). 6. What influences the probability of a Type II error? According to “Type I and Type II Errors” (n.d.): “The false negative rate is the proportion of positive instances that were erroneously reported as negative. It is equal to 1 minus the power of the test. False negative rate = number of false negatives/total number of positive instances. Type II errors can be caused by a lack of sensitivity or, In many cases, an oversight. 7. What is the difference between statistical significance and practical significance? Statistical significance means that the noticed mean variations are probably not due to a sampling error. Even a small sample, if it is large enough for the test, can work for statistical significance. Practical significance, on the other hand, considers if the difference is adequate enough to be of help in a practical sense. 8. Discuss the reasons that a researcher might obtain non-significant results. A researcher may wind up with non-significant results if the significance test demonstrates a high probability value. This means that the data cannot determine that the null hypothesis is false. On the other hand, this probability value still does not prove the truth of the null hypothesis. This often occurs because a researcher, in some cases, cannot determine a small effect from a null effect. (Interpreting Non-Significant Results, n.d.) Part II 1. Why would a researcher be concerned about generalizing to other subject populations? What are some of the subject population generalization problems that a researcher might confront? A researcher might be concerned about generalizing to other subject populations because populations can show significant variance, and a blanket generalization may, in fact, be untrue. This could be a problem for any researcher that is trying to prove a result to a test. Any researcher, for any research project, will probably be criticized about time, sample, and size in relation to the experiment. In a perfect world, a researcher would be able to get a sample equivalent to the population each time; however, it is not possible in this world. Therefore, researchers are forced to use generalization in order to obtain results. Generalization can hopefully help assist the researcher to select a sample size the closely or nearly represents the population in general. Of course, the closer, the better. Another item the researcher could consider, if using a smaller sample size, is to actually admit that the sample does not represent the population in general. However, this will all depend on what the researcher hopes to discover. 2. What is the source of the problem of generalizing to other experimenters? How can this problem be solved? Generalizing to other experiments can be a problem because every time an experiment is performed, different factors are involved. For example, to assume that I will get the same results as Student A because we are both doing the same study may not be a true assumption. Depending on the population sample and sometimes the time of the test (since attitudes and behaviors change sometimes from moment to moment) the sample size could be affected and our results could be different. Also, environments can sometimes be an issue. If Student A does his or her experiment in some bright, cheery area and I do mine in a room with blank walls and no windows, we are probably going to get some kind of different result, depending on what the test is. Therefore, so many factors can come into play that this can become a real problem. However, the problem can be solved. First, researchers need to admit to any differences in the studies that may affect the results differently. If that can be addressed, then the probability of generalizing can be established, and if the probability of the two studies being similar seems good, then a generalization can be made or assumed. 3. Why is it important to pretest a problem for generalization? Discuss the reasons why including a pretest may affect the ability to generalize results. It is important to pretest a problem for generalization to check on the sample size. Pretesting a problem will let a researcher see if he or she is actually getting a good result from the group, and he or she can also ensure that the sample itself is appropriate. Furthermore, pretests are always a good idea in order to avoid other issues that may hinder the study—behavioral issues, a depressing environment, or even a bad day; all of this can affect research when dealing with human subjects. The researcher can use a pretest to double check for any errors that may also be apparent in the study itself, especially if it is a survey, and also make sure that the data can be proven. A pretest may affect the ability to generalize results because it should indicate to the researcher whether there is anything wrong in the current generalization. If there is, then the researcher will need to reconsider certain areas of the study. 4. Distinguish between an exact replication and a conceptual replication. What is the value of a conceptual replication? Conceptual replication occurs when more than one observation that is relevant to a given relationship is made on each subject. This concept includes replicated measurement. Exact replication occurs when the observations to a given relationship made on each subject are exact. Conceptual replication is important because it still allows the researchers to view similarities. It is also much easier for researchers to attain conceptual replications, where the results are similar, rather than achieving an exact replication. 5. What is a meta-analysis? Meta-analysis can be used to summarize, amalgamate, and review quantitative research. In itself, it is a statistical technique, and it can assist with several different questions and concepts, as long as there is enough primary research studies to refer to. The researcher then can pick areas of the reported results of those studies, enter this information into a database, and then the information is “meta analyzed” to test other hypothesis ideas the researcher may have (Null, 2005). References Interpreting Non-Significant Results (n.d). Retrieved June 25, 2009, from: http://onlinestatbook.com/chapter9/nonsignificant.html Null, J (2005). Meta-analysis. Retrieved June 25, 2009, from: http://wilderdom.com/research/meta-analysis.html Null Hypothesis experiment resources. (n.d.) Retrieved June 25, 2009, from: http://www.experiment-resources.com/null-hypothesis.html 2008 Statistical Significance (1997). Stat Pack. Retrieved June 25, 2009, from: http://www.statpac.com/surveys/statistical-significance.htm. Type I and Type II errors (n.d). Retrieved June 25, 2009, from: http://davidmlane.com/hyperstat/A18652.html Read More