$FL2@(#) SPSS DATA FILE MS Windows 17.0.0 ################(#########Y@04 Oct 0916:49:14 ###########################PARTICIP########################CREATIVI####Creativit y pre-test ########################V3_A ####Creativity posttest############################################################## ############ ########################################### #######?###PARTICIP=Participant CREATIVI=CreativityPre V3_A=CreativityPost########################(################### ###windows1252#######efghxijklmnopqrstuvwxyz{ |}~}~ ~ $FL2@(#) SPSS DATA FILE MS Windows 17.0.0 ##########################Y@04 Oct 0921:13:01 ###########################SETTING ####Setting ########################SYSTOLIC####Systolic Blood Pressure ########################DIASTOLI####Diastolic Blood Pressure##############? #Home (control) #######@#Doctor's office#######@ Classroom ########################################################################## ############ ##################################################8###SETTING=Setting SYSTOLIC=SystolicBP DIASTOLI=DiastolicBP############################################ ###windows-1252#######e ee  e eeeeeeffffffffffgg gg g ggggg ###### NORTHCENTRAL UNIVERSITY ASSIGNMENT COVER SHEET Student: Stefanie L. McPherson THIS FORM MUST BE COMPLETELY FILLED IN Follow these procedures: If requested by your instructor, please include an assignment cover sheet. This will become the first page of your assignment. In addition, your assignment header should include your last name, first initial, course code, dash, and assignment number. This should be left justified, with the page number right justified. For example: DoeJXXX0000-1 1 Save a copy of your assignments: You may need to re-submit an assignment at your instructor's request. Make sure you save your files in accessible location. Academic integrity: All work submitted in each course must be your own original work. This includes all assignments, exams, term papers, and other projects required by your instructor. Knowingly submitting another person's work as your own, without properly citing the source of the work, is considered plagiarism. This will result in an unsatisfactory grade for the work submitted or for the entire course. It may also result in academic dismissal from the University. BTM8107-8 Dr. Yan Liu Statistics II Assignment #4 Faculty Use Only McPhersonSBTM8107-8-4 2 Part A. Dependent t test For this assignment, we are interested in finding out whether participation in a creative writing course results in increased scores of a creativity assessment. For this part of the activity, you will be using the data file \"Activity 4a.sav\". In this file, \"Participant\" is the numeric student identifier, \"CreativityPre\" contains creativity pre-test scores, and \"CreativityPost\" contains creativity post-test scores. A total of 40 students completed the pre-test, took the creativity course, and then took the post-test. 1. Exploratory Data Analysis/Hypotheses Here the null hypothesis H0: the participation in a creative writing course results is not increased in scores of creative assessment against the alternative H1: the participation in a creative writing course results is increased in scores of creative assessment. In other words we can say the null hypothesis H0: d=0 against the alternative hypothesis H1: d>0 where d is mean of the sample differences of the two samples. 2. Comparison of Means This is a paired sample t test. And the test statistics is: t = (differences of the observation) / ((N(differences of observation)2 - ((differences of observation))2) / (N-1)). Where N is the sample size, here N=40 The SPSS result: Paired Samples Statistics Mean Pair 1 N Std. Deviation Std. Error Mean Creativity pre-test 40.15 40 8.304 1.313 Creativity post-test 43.35 40 9.598 1.518 The means of Creativity pre-test is 40.15, standard deviation is 8.304 and the mean of Creativity post-test is 43.45 and standard deviation is 9.598. McPhersonSBTM8107-8-4 3 Paired Samples Correlations N Pair 1 Creativity pre-test & Correlation 40 .650 Sig. .000 Creativity post-test From the correlation matrix of these two variables we see that the correlation between them is 0.650 and also from the scatter plot we observe that the correlation between them is high and they are highly correlated. McPhersonSBTM8107-8-4 4 From the result we can see that the t value is -2.671, corresponding to the t value, p value is 0.011 which is less than 0.05 at 5% level of significance therefore we reject the null hypothesis and accept the alternative hypothesis that the participation in a creative writing course results is increased in scores of creative assessment. Part B. Independent t test We will start with the data file used in Part A (\"Activity 4a.sav\"). Suppose, however, you [the researcher] encountered a small problem during data collection: after the post-tests were collected, you realized that the post-test form did not ask for the students' identification number. As such, it will be impossible to match pre-test scores to post-test scores. Rather than simply give up, you start thinking about the data you do have and try to determine whether you can salvage your project. In assessing the situation, you realize that you have 40 pre-test scores and 40 post-test scores, but no way to link them. While it will result in a weaker comparison, you determine that you are still able to compare pre-test vs. post-test scores; you will use a betweensubjects design rather than a within-subjects design. 1. Create the data set. We arrange the data in such a manner that we change the variable by test and it is denoted by 1 for pre-test and 2 for post-test and combined sample as one of them. 2. Exploratory Data Analysis/Hypotheses. a) There is no difference between pre-test scores and post test scores Group Statistics Participant Creativity pre-test N Mean Std. Deviation Std. Error Mean 1 40 38.68 8.775 1.387 2 40 44.55 8.370 1.323 From the above we see that the mean of pre-test is 38.68 with standard deviation 8.775, the mean of post-test is 44.55 with standard deviation of 8.370. McPhersonSBTM8107-8-4 5 Here the null hypothesis H0: There is no difference between pre-test scores and post-test scores against the alternative H1: There is no difference between pre-test scores and posttest scores. 3. Comparison of Means Assuming equal or not equal variances both the cases corresponding to the t value the p value is 0.003 which is less than 0.05 at 5% level of significance therefore we reject the null hypothesis and accept the alternative hypothesis that there is a difference between pre and post-test scores. 4. Comparison of Designs a. You used the same dataset to analyze both a between- and within-subjects design. Create a single paragraph (using the material you wrote above), that presents both sets of results. The null hypothesis H0: there is no difference between pre-test scores and post-test scores; against the alternative H1: there is no difference between pre-test scores and posttest scores. Assuming equal or not equal variances both the cases corresponding to the t value the p value is 0.003 which is less than 0.05 at 5% level of significance therefore we reject the null hypothesis and accept the alternative hypothesis that there is a difference between pre and post-test scores. b. Explain, in 300-500 words, whether the two tests resulted in the same findings. Did you expect this to be the case? Why or why not? What have you learned in this assignment? Since there was no difference between the pre-test and post-test scores, yes, the two tests resulted in the same findings which was expected. Between subject designs are very valuable because they provide opportunities for the researcher to conduct an experiment McPhersonSBTM8107-8-4 6 without extraneous factors. This design is really considered independent because every participant completes a pre-test and post-test only, there aren't any other tests skewing the results. There are disadvantages with \"between subject design\" such as, if the experiment is rare and the researcher has a difficult time finding candidates to participate. Another variability to \"between subjects design\" is taking into account an individual's emotional response to test taking, some people get nervous and will score lower because of that response. It is important to take those variability's into account when designing the experiment. Also, it is important for the researcher to be unbiased when selecting candidates for the study, do not select individuals based on their intelligence or ability to achieve a high score on a test, this will skew the results. A researcher should use a random selection process for the research in order to achieve true results. Also, a researcher should take into account any environmental factors, such as administering all the tests in the morning when individuals are at their mental peak. The study should eliminate as much bias as possible. This experiment is a \"within subject design\" because each participant performed a pretest and a post-test. In a \"within subject design\" experiment each participant is subjected to every single treatment. The main advantage of a \"within subject design\" is that it requires fewer participants, and is more streamlined because it is not resource dependent. A disadvantage with this type of research design is that sometimes the first test influences the second test. I believe that was the case with this experiment because the results were the same. Also, the researcher failed to link the scores, but since the results were the same, perhaps that was not necessary. Part C. ANOVA All of us have had our blood pressure measured while at our physician's office. How accurate are these measurements? It may surprise you to learn that there is something called \"white coat syndrome\"the tendency of some people to exhibit elevated blood pressure in clinical (medical) settings only. In other words, for these people, the very fact that the physician is taking their blood pressure causes it to increase In this activity, you will be using the \"Activity 4c.sav\" data file to determine whether you find support for the existence of white coat syndrome. In this study, 60 participants were randomly assigned to one of three groups. The \"settings\" variable indicates the location in which the participants' blood pressure was recorded: 1=home, 2=in a doctor's office, and 3=in a classroom setting. The \"SystolicBP\" variable contains the participants' systolic pressure (the \"upper\" number). The \"DiastolicBP\" variable contains the participant's diastolic pressure (the \"lower\" number). 1. Exploratory Data Analysis/Hypotheses. a. Perform exploratory data analysis on both the SystolicBP and DiastolicBP variables. Using SPSS, calculate the mean and standard deviation of these two variables. Be sure that your analysis is broken down by setting (e.g., you will have six means, six SD's, etc.). McPhersonSBTM8107-8-4 The outputs of SPSS containing the mean and standard deviation of the three levels of these two variables are shown below: 7 b. Create two graphsone for systolic and one for diastolic pressure. Each graph should clearly delineate the three groups. The comparative box-plots for the three levels of these two variables drawn using SPSS is shown below: McPhersonSBTM8107-8-4 The box-plots depict that the distributions of all the three levels of the variables \"Systolic Blood Pressure\" and \"Diastolic Blood Pressure\" are approximately normal. c. Write a null and alternative hypothesis for the comparison of the three groups (note that your hypothesis will state that the three groups are equivalent; be sure to word your null hypothesis correctly). The null and alternative hypotheses for systolic blood pressure are, H0: There is no significant difference between the mean systolic blood pressures of the three levels, 1 = 2 = 3. 8 McPhersonSBTM8107-8-4 Ha: At least two of the three levels differ significantly in mean systolic blood pressures. 9 The null and alternative hypotheses for diastolic blood pressure are, H0: There is no significant difference between the mean diastolic blood pressures of the three levels, 1 = 2 = 3. Ha: At least two of the three levels differ significantly in mean diastolic blood pressures. 2. ANOVA. a. Using the \"Activity 4c.sav\" data file, perform two single factor ANOVAs: one using SystolicBP and one using DiastolicBP as the dependent variable. The output of SPSS containing the results of one-way ANOVA for the two variables are shown below: McPhersonSBTM8107-8-4 10 b. If appropriate for either or both of the ANOVAs, perform post hoc analyses to determine which groups actually differ. The output of SPSS containing the results of post-hoc tests on the three levels of the variable \"Systolic Blood Pressure\" is shown below: c. Write one paragraph for each ANOVA (be sure to use APA style). At a minimum, each paragraph should contain the three means, three SD's, ANOVA results (F, df), post hoc tests (if applicable), effect size, and an interpretation of these results. Systolic Blood Pressure: The results of one-way ANOVA to test for Systolic Blood Pressure differences among three levels indicated that Systolic Blood Pressure differed significantly across the three levels, F (2,27) = 9.964, p = .001 < 0.005. McPhersonSBTM8107-8-4 11 The effect size for this ANOVA is 0.425. It is moderately high. The results of Tukey's post-hoc comparisons of the three levels indicated that the level Doctor's office (M = 132.60, 95% CI [126.61, 138.59]) gave significantly larger scores than the levels Home (M = 122.90, 95% CI [117.83, 127.97]), p = 0.13 and Class room (M = 118.80, 95% CI [114.83, 122.77]), p = .001. Comparison between the levels' Home and Class room (p = 0.412) is not significant. Diastolic Blood Pressure: The results of one-way ANOVA to test for Diastolic Blood Pressure differences among three levels indicated that Diastolic Blood Pressure does not differ significantly across the three levels, F (2, 27) = 0.105, p = .900 > 0.10. The effect size for this ANOVA is 0.008. It is significantly very small. NORTHCENTRAL UNIVERSITY ASSIGNMENT COVER SHEET Student: Stefanie L. McPherson THIS FORM MUST BE COMPLETELY FILLED IN Follow these procedures: If requested by your instructor, please include an assignment cover sheet. This will become the first page of your assignment. In addition, your assignment header should include your last name, first initial, course code, dash, and assignment number. This should be left justified, with the page number right justified. For example: DoeJXXX0000-1 1 Save a copy of your assignments: You may need to re-submit an assignment at your instructor's request. Make sure you save your files in accessible location. Academic integrity: All work submitted in each course must be your own original work. This includes all assignments, exams, term papers, and other projects required by your instructor. Knowingly submitting another person's work as your own, without properly citing the source of the work, is considered plagiarism. This will result in an unsatisfactory grade for the work submitted or for the entire course. It may also result in academic dismissal from the University. BTM8107-8 Dr. Yan Liu Statistics II Assignment #4 Faculty Use Only McPhersonSBTM8107-8-4 2 Part A. Dependent t test For this assignment, we are interested in finding out whether participation in a creative writing course results in increased scores of a creativity assessment. For this part of the activity, you will be using the data file \"Activity 4a.sav\". In this file, \"Participant\" is the numeric student identifier, \"CreativityPre\" contains creativity pre-test scores, and \"CreativityPost\" contains creativity post-test scores. A total of 40 students completed the pre-test, took the creativity course, and then took the post-test. 1. Exploratory Data Analysis/Hypotheses Here the null hypothesis H0: the participation in a creative writing course results is not increased in scores of creative assessment against the alternative H1: the participation in a creative writing course results is increased in scores of creative assessment. In other words we can say the null hypothesis H0: d=0 against the alternative hypothesis H1: d>0 where d is mean of the sample differences of the two samples. 2. Comparison of Means This is a paired sample t test. And the test statistics is: t = (differences of the observation) / ((N(differences of observation)2 - ((differences of observation))2) / (N-1)). Where N is the sample size, here N=40 The SPSS result: Paired Samples Statistics Mean Pair 1 N Std. Deviation Std. Error Mean Creativity pre-test 40.15 40 8.304 1.313 Creativity post-test 43.35 40 9.598 1.518 The means of Creativity pre-test is 40.15, standard deviation is 8.304 and the mean of Creativity post-test is 43.45 and standard deviation is 9.598. McPhersonSBTM8107-8-4 3 Paired Samples Correlations N Pair 1 Creativity pre-test & Correlation 40 .650 Sig. .000 Creativity post-test From the correlation matrix of these two variables we see that the correlation between them is 0.650 and also from the scatter plot we observe that the correlation between them is high and they are highly correlated. McPhersonSBTM8107-8-4 4 From the result we can see that the t value is -2.671, corresponding to the t value, p value is 0.011 which is less than 0.05 at 5% level of significance therefore we reject the null hypothesis and accept the alternative hypothesis that the participation in a creative writing course results is increased in scores of creative assessment. Part B. Independent t test We will start with the data file used in Part A (\"Activity 4a.sav\"). Suppose, however, you [the researcher] encountered a small problem during data collection: after the post-tests were collected, you realized that the post-test form did not ask for the students' identification number. As such, it will be impossible to match pre-test scores to post-test scores. Rather than simply give up, you start thinking about the data you do have and try to determine whether you can salvage your project. In assessing the situation, you realize that you have 40 pre-test scores and 40 post-test scores, but no way to link them. While it will result in a weaker comparison, you determine that you are still able to compare pre-test vs. post-test scores; you will use a betweensubjects design rather than a within-subjects design. 1. Create the data set. We arrange the data in such a manner that we change the variable by test and it is denoted by 1 for pre-test and 2 for post-test and combined sample as one of them. 2. Exploratory Data Analysis/Hypotheses. a) There is no difference between pre-test scores and post test scores Group Statistics Participant Creativity pre-test N Mean Std. Deviation Std. Error Mean 1 40 38.68 8.775 1.387 2 40 44.55 8.370 1.323 From the above we see that the mean of pre-test is 38.68 with standard deviation 8.775, the mean of post-test is 44.55 with standard deviation of 8.370. McPhersonSBTM8107-8-4 5 Here the null hypothesis H0: There is no difference between pre-test scores and post-test scores against the alternative H1: There is no difference between pre-test scores and posttest scores. 3. Comparison of Means Assuming equal or not equal variances both the cases corresponding to the t value the p value is 0.003 which is less than 0.05 at 5% level of significance therefore we reject the null hypothesis and accept the alternative hypothesis that there is a difference between pre and post-test scores. 4. Comparison of Designs a. You used the same dataset to analyze both a between- and within-subjects design. Create a single paragraph (using the material you wrote above), that presents both sets of results. The null hypothesis H0: there is no difference between pre-test scores and post-test scores; against the alternative H1: there is no difference between pre-test scores and posttest scores. Assuming equal or not equal variances both the cases corresponding to the t value the p value is 0.003 which is less than 0.05 at 5% level of significance therefore we reject the null hypothesis and accept the alternative hypothesis that there is a difference between pre and post-test scores. b. Explain, in 300-500 words, whether the two tests resulted in the same findings. Did you expect this to be the case? Why or why not? What have you learned in this assignment? Since there was no difference between the pre-test and post-test scores, yes, the two tests resulted in the same findings which was expected. Between subject designs are very valuable because they provide opportunities for the researcher to conduct an experiment McPhersonSBTM8107-8-4 6 without extraneous factors. This design is really considered independent because every participant completes a pre-test and post-test only, there aren't any other tests skewing the results. There are disadvantages with \"between subject design\" such as, if the experiment is rare and the researcher has a difficult time finding candidates to participate. Variability to \"between subjects design\" is taking into account an individual's emotional response to test taking, some people get nervous and will score lower because of that response. It is important to take those variability's into account when designing the experiment. Also, it is important for the researcher to be unbiased when selecting candidates for the study, do not select individuals based on their intelligence or ability to achieve a high score on a test, this will skew the results. A researcher should use a random selection process for the research in order to achieve true results. Also, a researcher should take into account any environmental factors, such as administering all the tests in the morning when individuals are at their mental peak. The study should eliminate as much bias as possible. This experiment is a \"within subject design\" because each participant performed a pretest and a post-test. In a \"within subject design\" experiment each participant is subjected to every single treatment. The main advantage of a \"within subject design\" is that it requires fewer participants, and is more streamlined because it is not resource dependent. A disadvantage with this type of research design is that sometimes the first test influences the second test. I believe that was the case with this experiment because the results were the same. Also, the researcher failed to link the scores, but since the results were the same, perhaps that was not necessary. Part C. ANOVA All of us have had our blood pressure measured while at our physician's office. How accurate are these measurements? It may surprise you to learn that there is something called \"white coat syndrome\"the tendency of some people to exhibit elevated blood pressure in clinical (medical) settings only. In other words, for these people, the very fact that the physician is taking their blood pressure causes it to increase In this activity, you will be using the \"Activity 4c.sav\" data file to determine whether you find support for the existence of white coat syndrome. In this study, 60 participants were randomly assigned to one of three groups. The \"settings\" variable indicates the location in which the participants' blood pressure was recorded: 1=home, 2=in a doctor's office, and 3=in a classroom setting. The \"SystolicBP\" variable contains the participants' systolic pressure (the \"upper\" number). The \"DiastolicBP\" variable contains the participant's diastolic pressure (the \"lower\" number). 1. Exploratory Data Analysis/Hypotheses. a. Perform exploratory data analysis on both the SystolicBP and DiastolicBP variables. Using SPSS, calculate the mean and standard deviation of these two variables. Be sure that your analysis is broken down by setting (e.g., you will have six means, six SD's, etc.). McPhersonSBTM8107-8-4 The outputs of SPSS containing the mean and standard deviation of the three levels of these two variables are shown below: 7 b. Create two graphsone for systolic and one for diastolic pressure. Each graph should clearly delineate the three groups. The comparative box-plots for the three levels of these two variables drawn using SPSS is shown below: McPhersonSBTM8107-8-4 The box-plots depict that the distributions of all the three levels of the variables \"Systolic Blood Pressure\" and \"Diastolic Blood Pressure\" are approximately normal. c. Write a null and alternative hypothesis for the comparison of the three groups (note that your hypothesis will state that the three groups are equivalent; be sure to word your null hypothesis correctly). The null and alternative hypotheses for systolic blood pressure are, H0: There is no significant difference between the mean systolic blood pressures of the three levels, 1 = 2 = 3. 8 McPhersonSBTM8107-8-4 Ha: At least two of the three levels differ significantly in mean systolic blood pressures. 9 The null and alternative hypotheses for diastolic blood pressure are, H0: There is no significant difference between the mean diastolic blood pressures of the three levels, 1 = 2 = 3. Ha: At least two of the three levels differ significantly in mean diastolic blood pressures. 2. ANOVA. a. Using the \"Activity 4c.sav\" data file, perform two single factor ANOVAs: one using SystolicBP and one using DiastolicBP as the dependent variable. The output of SPSS containing the results of one-way ANOVA for the two variables are shown below: McPhersonSBTM8107-8-4 10 b. If appropriate for either or both of the ANOVAs, perform post hoc analyses to determine which groups actually differ. The output of SPSS containing the results of post-hoc tests on the three levels of the variable \"Systolic Blood Pressure\" is shown below: c. Write one paragraph for each ANOVA (be sure to use APA style). At a minimum, each paragraph should contain the three means, three SD's, ANOVA results (F, df), post hoc tests (if applicable), effect size, and an interpretation of these results. Systolic Blood Pressure: The results of one-way ANOVA to test for Systolic Blood Pressure differences among three levels indicated that Systolic Blood Pressure differed significantly across the three levels, F (2,27) = 9.964, p = .001 < 0.005. McPhersonSBTM8107-8-4 11 The effect size for this ANOVA is 0.425. It is moderately high. The results of Tukey's post-hoc comparisons of the three levels indicated that the level Doctor's office (M = 132.60, 95% CI [126.61, 138.59]) gave significantly larger scores than the levels Home (M = 122.90, 95% CI [117.83, 127.97]), p = 0.13 and Class room (M = 118.80, 95% CI [114.83, 122.77]), p = .001. Comparison between the levels' Home and Class room (p = 0.412) is not significant. Diastolic Blood Pressure: The results of one-way ANOVA to test for Diastolic Blood Pressure differences among three levels indicated that Diastolic Blood Pressure does not differ significantly across the three levels, F (2, 27) = 0.105, p = .900 > 0.10. The effect size for this ANOVA is 0.008. It is significantly very small