Write a paper on Correlation Analysis

Write a paper on Correlation AnalysisU‌‍‍‍‌‍‍‌‌‌‍‌‍‍‍‍‌‍‌‌sing the demographic variables from Assignment order 82680275 For each of these variables you will conduct the appropriate correlational analysis with at least one other variable from the Programs Outcomes Data Base that is related to your specific program. It is recommended that you conduct correlational analysis on two variables that are at the interval level or higher so that you can just interpret the Pearson’s r measure of association. Using APA format for reporting statistics, report the results of your correlations. Then provide the interpretation and conclusion based on the strength of the relationship. Attach a copy of the actual analysis (EXCEL Spreadsheet file) to the assignment. https://youtu.be/fxRyunAQR48 To activate it. 1. To install Excel’s Analysis Toopak, click the File tab on the top-left and then click Options on the bottom-left. 2. Then, click Add-Ins. On the Manage drop-down list, choose Excel Add-ins, and click GO. On the popup that appears, check Analysis ToolPak and click OK. 3. After you enable it, click Data Analysis in the Data menu to display the analyses you can perform. References Ham, ., Huggins-Hoyt, Y. & Pettuss, J. (2016). Assessing statistical change indices in selected social work intervention research studies. Research on Social Work Practice, 26(1), 44-52. Rubin, A. (2010). Statistics for evidence-based practice and evaluation, (3rd ed.). Brooks Cole. Please read the enclosed information and answer the following questions The Process of Hypothesis Testing When inferential statistics are used as part of the analytic plan in evaluation, the intent is to remove chance as an explanation for the result. It’s important to recognize that when a hypothesis is developed, from an analysis point of view, this is referred to as an alternate hypothesis. However, with each alternate hypothesis there is a null hypothesis. You are using your data to test the null hypothesis. Regardless of the inferential procedure, the steps in hypothesis testing are the same. The hypothesis is developed, the necessary statistical procedures identified, and the significance value or probability value is determined that will eventually used to determine if the null hypothesis is rejected or accepted. The nature of the hypothesis will also have implications for whether a one-tailed test or two-tailed test will be used. Generally speaking, if the alternate hypothesis provides a direction to the relationship, such as increase or decrease, a one tailed test can be used. If the hypothesis is that there are only differences but no direction, a two-tailed test would be indicated for the interpretation of the test data. The data has been collected and the statistical analysis is completed. The analysis yields a test statistic. The test statistic yielded by the statistical test is used for comparison with the critical value associated with the significance level that you have identified at the beginning of the process. It’s important to note however that statistical software packages, including EXCEL, will provide the actual significance level of the result obtained. In this case, you just compare the p-value in your results with the p-value you establish at the beginning of the study to make judgments about the data. Make the final interpretation of the results. As noted above, you compare the p value against the p value you established at the beginning of the study. If the p-value associated with the results is smaller than the p-value you established as your criterion, you reject the null hypothesis. Validity, Once Again! During your studies in research methods you may have noticed this word “validity” comes up quite often. You’ve previous seen the term used in conjunction with measurement (Construct, Content, and Criterion Validity); again, in discussing the integrity of various experimental designs (internal and external validity); and once again, here, as it relates to statistical judgments – statistical conclusion validity. As the other validities were concerned with some aspect of accuracy, this is true as it relates to the interpretation of statistical results. The discussion of statistical conclusion validity is framed in terms of the risk for Type I versus Type II error in accepting or rejecting the null hypothesis. Your text provides a thorough discussion of the relevance of these types of error in making judgments. This issue will also become relevant in the discussion on calculating sample size in a couple of weeks. Correlation and Correlational Design‌‍‍‍‌‍‍‌‌‌‍‌‍‍‍‍‌‍‌‌s Correlation refers to “the degree to which the values of two variables vary together in a consistent fashion.” (Rubin, , 2013), Depending on the strength of the consistency it may be reasonable to assume that there is some type of relationship among the two variables, but, due to the limitations of the way the data is collected and analyzed, it cannot be said that one variable is causing the change in the other variable. In fact, another variable, outside of the scope of observation of the study or evaluation could be accounting for the change in value of both variables. There are numerous tests of correlation, that as you have learned, exist due to the varying level of measurement of either of the two variables. For instance, if both variables are measured at the interval or ratio level, the Pearson’s product-moment correlation is the proper statistic to be used. Spearman’s rho is a test of correlation when the variables are at the ordinal level. In the case of phi coefficient and Cramer’s V are preferred tests when both variables are at the nominal level of measurement. Regardless of the test, all yield a result between -1 and 1, which is an indication of the strength of the relationship. As with all inferential statistics a p-value is calculated. An extension of correlation analyses is regression. If there is a strong correlation among variables, the value of one variable could be used to predict the value of another variable. This would be a bivariate analysis, known as single regression analysis. However, regression can be used with many independent variables (predictors) on one dependent (response) variable, this is known as multiple regression analysis. As with all other statistical procedures, the level of measurement of the variables plays a role in determining the actual regression analysis. Linear regression is used when both the independent (predictor) variable are continuous (interval, ratio level). Dummy coding can be used on categorical variables for use with linear regression. When the dependent variable is categorical, logistic regression is the appropriate procedure. Test Your Knowledge Testing and Measuring Relationships For each of the following scenarios, please indicate the correct answer. With most measures of association a perfect relationship is indicated as: .05 .001 .80 Which of the following statements is true about significance levels? They indicate how strong a relationship is. A relationship significant at the .001 level is stronger than one significant at the .05 level of significance. A weak relationship can be significant at the .001 level if the sample size is very large. A strong relationship will always be significant at the .05 level, regardless of the sample size. The .05 level of significance is commonly chosen because: Mathematics dictates it as the only correct level. It means there is at least a .05 correlation between the independent and dependent variables. Two groups would have to show a .05 difference to be significant. None of the above. Statistical significance refers to: The extent to which your method of measurement is reliable and valid. The extent to which you can rule out chance as the explanation of your clients’ measured growth rather than the intervention. The extent to which you can generalize your findings to persons not included in your study. The extent to which you can replicate your study with other persons. Type I and Type II error affect: Critical value Statistical conclusion validity Internal validity Significance Level Interpreting correlation or measures of association, which of the following is indicates the strongest correlation? .56 .05 .86 10 Submit Reset Summary Module three has focused on the first of the statistical procedures that have employed the principles of hypothesis testing. The process of hypothesis testing will be the same for all inferential statistics. The null and hypotheses are developed. The researcher selects the p-value which serves as the basis for the conclusions on the null hypothesis. The data is collected, and the statistical procedure is performed. The results are reviewed and the conclusions on the null hypothesis are made. Reading Resources Rubin, A. (2013). Statistics for Evidence-Based practice and evaluation (3rd ed). Brooks Cole. Chapters 11- 13; 17 & 18 Ham, ., Huggins-Hoyt, Y. & Pettuss, J. (2016). Assessing statistical change indices in selected social work intervention research studies. Research on Social Work Practice, 26(1), 4‌‍‍‍‌‍‍‌‌‌‍‌‍‍‍‍‌‍‌‌4-52.