Darius Muniz
Professor Godfrey
Research Methods II
September 19th, 2011
SPSS TextBook
I. a) HeartRate Scale Ratio
b) Nervous Scale Interval
c) Sex Nominal Nominal
d) YearsInCollege Ordinal Ordinal
e) ExperimentalGroup Nominal Nominal
II. Nominal and Ordinal Variables:
What year are you in college?  
 Frequency  Percent  Valid Percent  Cumulative Percent  
Valid  Junior  17  44.7  51.5  51.5 
Senior  16  42.1  48.5  100.0  
Total  33  86.8  100.0 
 
Missing  99  5  13.2 


Total  38  100.0 

 





A) For “what year in college” (an ordinal variable) out of 38 students surveyed, 51 percent reported being in their junior year of college, 49 percent reported being in their senior year, and no students identified as freshman or sophomore.
B) 33 students, which made up 87 percent of the population, reported their academic level with 17 self identifying as juniors (representing 45 percent of the class) and 16 self identifying as seniors (representing 42 percent of the class). 5 students did not answer the question (representing 13 percent of the class) for a total population of N=38 students.
What is your sex  
 Frequency  Percent  Valid Percent  Cumulative Percent  
Valid  Male  7  18.4  21.2  21.2 
Female  26  68.4  78.8  100.0  
Total  33  86.8  100.0 
 
Missing  99  5  13.2 


Total  38  100.0 


A) For “what is your sex” (a nominal variable) out of 38 students surveyed, 21percent self identified as male and 79 percent self identified as female.
B) 33 students, which made up 87 percent of the population, reported their sex with 7 students (representing 18 percent of the class) self identifying as male and 26 students (representing 69 percent of the class) self identifying as female. 5 students did not answer the question (representing 13 percent of the class) for a total population of N=38 students.
What experimental group were you in?  
 Frequency  Percent  Valid Percent  Cumulative Percent  
Valid  A (something good that happened to you)  9  23.7  27.3  27.3 
B(something good happened to someone you knew)  10  26.3  30.3  57.6  
C(write letters)  14  36.8  42.4  100.0  
Total  33  86.8  100.0 
 
Missing  99  5  13.2 


Total  38  100.0 


A) For “what experimental group were you in” (a nominal variable) out of 38 students surveyed, 27 percent were assigned to condition A, 30 percent were assigned to condition B, and 42 percent were assigned to condition C.
B) 33 students, which made up 87 percent of the population, reported which group they were assigned to with 9 students in condition A (representing 24 percent of the population), 10 students in condition B (representing 26 percent of the population), and 14 students in condition C (representing 37 percent of the population). 5 students did not answer the question (representing 13 percent of the population) for a total population of N=38 students.
III. HeartRate and Cheerful:
Statistics  
 Heart Rate  Right now, I feel cheerful...  
N  Valid  38  38 
Missing  0  0  
Mean  65.79  2.89  
Std. Deviation  14.994  1.060  
Minimum  10  1  
Maximum  90  5 
A) The interval variable, “right now I feel cheerful”, had a mean of 2.89 and a standard deviation of 1.060. The interval variable, “HeartRate”, had a mean of 65.79 and a standard deviation of 14.994.
B) Out of 38 students the average (mean) heart rate reported was around 66 beats per minute with some students falling either 15 beats below the mean or 15 beats above the mean. The average response from students when asked to rate how cheerful they felt right now was 3 (somewhat) with some students rating themselves below the mean as 2 (a little) or above the mean as 4 (a lot).
C) The histogram of the variable “right now I feel cheerful” displays no outliers and would be considered a normal distribution with most scores tightly clustered around the mean.
The histogram of the variable “HeartRate” displays the presence of outliers and would be considered a negatively skewed distribution with the outliers pulling the tail towards the left of the graph away from the mean.
Darius Muniz
Professor Godfrey
Research Methods II
September 26th, 2011
TTest
A. The Null hypothesis is that there is no difference whether being in group A (writing something good that happened to you) or being in group C (writing the alphabet) and which neither group has an effect on heart Rate.
The Alternate/Experimental hypothesis is that being in group A or group C does make a difference and that being in a specific group did have an effect on Heart Rate.
B.
TTEST GROUPS=ExperimentalGroup(1 3)
/MISSING=ANALYSIS
/VARIABLES=HeartRate
/CRITERIA=CI(.95).
Group Statistics  
 What experimental group were you in?  N  Mean  Std. Deviation  Std. Error Mean 
Heart Rate  A (something good that happened to you)  9  64.67  12.649  4.216 
C(write letters)  14  64.71  11.809  3.156 
Independent Samples Test  
 Levene's Test for Equality of Variances  ttest for Equality of Means  
F  Sig.  t  df  Sig. (2tailed)  Mean Difference  Std. Error Difference  95% Confidence Interval of the Difference  
Lower  Upper  
Heart Rate  Equal variances assumed  .089  .768  .009  21  .993  .048  5.185  10.830  10.735 
Equal variances not assumed 

 .009  16.323  .993  .048  5.267  11.195  11.099 
C. An independentsamples ttest was conducted to compare the heart rates for groups A (write 90 seconds about something good that happened to you) N=9 and group C (write the alphabet backwards for 90 seconds) N=14. There was no significant difference in scores of heart rates between group A, (M=64.67, SD=12.649) and group C, (M=64.71, SD=11.809); t(21)= 0.01, p=.99.
D. For this analysis, we were interested in knowing if whether being in group A (write 90 seconds about something good that happened to you) or group C (write the alphabet backwards for 90 seconds) had an effect on Heart Rate. Since the ttest was not significant, we can conclude that there was no difference in heart rate contributed to being in either group A or group B other than what would be expected to occur by chance. This study demonstrates that there is no association between experimental group/writing condition and HeartRate.
ANOVA
A. The Null hypothesis is that experimental group does not affect “being in good spirits”.
The Alternate/Experimental hypothesis is that experimental group does have an effect on being “in good spirits”.
B.
ONEWAY InGoodSpirits BY ExperimentalGroup
/MISSING ANALYSIS
/POSTHOC=TUKEY ALPHA(0.05).
ANOVA  
Right now, I feel in good spirits...  
 Sum of Squares  df  Mean Square  F  Sig. 
Between Groups  15.598  2  7.799  11.747  .000 
Within Groups  19.917  30  .664 


Total  35.515  32 



Multiple Comparisons  
Right now, I feel in good spirits... Tukey HSD  
(I) What experimental group were you in?  (J) What experimental group were you in?  Mean Difference (IJ)  Std. Error  Sig.  95% Confidence Interval  
Lower Bound  Upper Bound  
A (something good that happened to you)  B(something good happened to someone you knew)  .811  .374  .094  .11  1.73 
C(write letters)  .817  .348  .064  1.68  .04  
B(something good happened to someone you knew)  A (something good that happened to you)  .811  .374  .094  1.73  .11 
C(write letters)  1.629*  .337  .000  2.46  .80  
C(write letters)  A (something good that happened to you)  .817  .348  .064  .04  1.68 
B(something good happened to someone you knew)  1.629*  .337  .000  .80  2.46  
*. The mean difference is significant at the 0.05 level. 
Right now, I feel in good spirits...  
Tukey HSDa,b  
What experimental group were you in?  N  Subset for alpha = 0.05  
1  2  
B(something good happened to someone you knew)  10  2.30 

A (something good that happened to you)  9  3.11  3.11 
C(write letters)  14 
 3.93 
Sig. 
 .072  .069 
Means for groups in homogeneous subsets are displayed.  
a. Uses Harmonic Mean Sample Size = 10.618.  
b. The group sizes are unequal. The harmonic mean of the group sizes is used. Type I error levels are not guaranteed. 
C. A oneway analysis of variance (ANOVA) was performed to investigate if “being in good spirits” was affected by experimental group A (writing 90 seconds about something good that happened to you) or experimental group C (writing backwards for 90 seconds). The ANOVA was statistically significant, F(2,30)=11.75, p<.001. Followup tests were to evaluate differences in pairs between the means. Tukey’s HSD was used to conduct the post hoc comparisons. There was a significant difference between the means of writing condition B and writing condition C. The mean for experimental group being in writing condition C “writing letters” (M=3.93) is higher than the means for both writing condition B “write something good about someone else” (M= 3.11) and writing condition A “write something good about you” (M= 2.30).
D. We wanted to find out if students who felt in good spirits were affected by the writing condition they were assigned to. The results showed that there was a significant difference about being in good spirits between the group of students who were assigned to writing condition C and the group of students who were assigned to condition B. Students who were in writing condition C (wrote the alphabet) reported being in good spirits more than students who were in writing condition B(wrote about someone else), and even greater spirits than students who were in writing condition A (who wrote about themselves).
Darius Muniz
Professor Godfrey
Research Methods II
October 23, 2011
Analysis 1
Run bivariate correlations between quality of sleep, stress over the last month and anxiety.
A. Copy and paste the correlation matrix from the output below.
DATASET ACTIVATE DataSet1.
CORRELATIONS
/VARIABLES=stressmo qualslp anxiety
/PRINT=TWOTAIL NOSIG
/MISSING=PAIRWISE.
Correlations  
 how stressed over last month  quality of sleep  HADS Anxiety  
how stressed over last month  Pearson Correlation  1  .247**  .532** 
Sig. (2tailed) 
 .000  .000  
N  271  268  268  
quality of sleep  Pearson Correlation  .247**  1  .360** 
Sig. (2tailed)  .000 
 .000  
N  268  268  265  
HADS Anxiety  Pearson Correlation  .532**  .360**  1 
Sig. (2tailed)  .000  .000 
 
N  268  265  268  
**. Correlation is significant at the 0.01 level (2tailed). 
B. Write one paragraph of a Results section to report the findings in APA style. Be sure to include the significant and nonsignificant findings and the statistics.
In order to determine the correlation between quality of sleep, stress over last month and anxiety, a correlations analysis was performed. Anxiety was significantly positively correlated with stress over the last month (r = .53, p<.001) and negatively correlated with quality of sleep (r = .36, p<.001). Quality of sleep was significantly negatively correlated with stress over the last month (r =.25, p<.001).
Analysis 2
First, run a linear regression using sex to predict anxiety.
A. Copy and paste the output from the linear regression below. (This should include the model summary, ANOVA and coefficients boxes).
Model Summary  
Model  R  R Square  Adjusted R Square  Std. Error of the Estimate  
dimension0  1  .247a  .061  .057  3.390 
a. Predictors: (Constant), Sex  
ANOVAb  
Model  Sum of Squares  df  Mean Square  F  Sig.  
1  Regression  197.998  1  197.998  17.224  .000a 
Residual  3057.778  266  11.495 

 
Total  3255.776  267 


 
a. Predictors: (Constant), Sex  
b. Dependent Variable: HADS Anxiety 
Coefficientsa  
Model  Unstandardized Coefficients  Standardized Coefficients  t  Sig.  
B  Std. Error  Beta  
1  (Constant)  5.388  .308 
 17.482  .000 
Sex  1.727  .416  .247  4.150  .000  
a. Dependent Variable: HADS Anxiety 
B. After seeing the output from the regression analyses, in your own words, describe what each of the following pieces of output are and what are they mean for this example: R square, regression Ftest, the intercept and the coefficient (b) for sex.
1. R Square: 6.1% of the variance in anxiety can be attributed to sex.
2. Regression FTest: Is it large enough than what we could expect by chance alone, yes because 17.22, p<.001, which is significant.
3. Intercept and Coefficient (b) for sex: The intercept is 5.39 which is the predicted score on anxiety when sex=0, or when sex is male, the positive sign shows that as sex increases by one unit or changes to female, anxiety increases by 1.73, and Beta tells us that for every standard deviation increase or change from male to female, anxiety increases by .25 standard deviations.
C. Write one paragraph of a Results section to report the findings in APA style. Be sure to include the significant and nonsignificant findings and the statistics from each of the analyses.
Standard multiple regression was used to assess the ability of sex to predict anxiety. The total variance explained by the model was 6.1%, F (1,266) = 17.22, p < .001. For every unit of sex, anxiety increased by 1.73 (beta=.25, b = 1.73, p<.001), making sex a significant predictor for anxiety in this sample.
D. Suppose you want to explain the results of this analysis to a friend who is not in the class. Translate the results paragraph into a onesentence takehome message.
This analysis shows that when a person’s sex changes from male to female it is positively associated with the amount of anxiety that a person feels.
Analysis 3
First, recode quality of sleep such that it ranges from 0 to 5 and call the new variable “quality of sleep recoded”. We will do this together in lab.
Second, run a hierarchical linear regression using quality of sleep recoded and stress over last month to predict anxiety. In the first step, include quality of sleep recoded as the predictor. In the second step, add stress over last month as a second predictor. Make sure you check the “change in Rsquare” box under the statistics tab.
A. Copy and paste the output from the linear regression below. (This should include the model summary, ANOVA and coefficients boxes).
Model Summary  
Model  R  R Square  Adjusted R Square  Std. Error of the Estimate  Change Statistics  
R Square Change  F Change  df1  df2  Sig. F Change  
dimension0  1  .360a  .130  .127  3.247  .130  39.261  1  263  .000 
2  .579b  .335  .330  2.844  .205  80.844  1  262  .000  
a. Predictors: (Constant), QualslpRecoded  
b. Predictors: (Constant), QualslpRecoded, how stressed over last month 
ANOVAc  
Model  Sum of Squares  df  Mean Square  F  Sig.  
1  Regression  413.844  1  413.844  39.261  .000a 
Residual  2772.217  263  10.541 

 
Total  3186.060  264 


 
2  Regression  1067.543  2  533.771  66.012  .000b 
Residual  2118.518  262  8.086 

 
Total  3186.060  264 


 
a. Predictors: (Constant), QualslpRecoded  
b. Predictors: (Constant), QualslpRecoded, how stressed over last month  
c. Dependent Variable: HADS Anxiety 
Coefficientsa  
Model  Unstandardized Coefficients  Standardized Coefficients  t  Sig.  
B  Std. Error  Beta  
1  (Constant)  9.307  .513 
 18.151  .000 
QualslpRecoded  1.097  .175  .360  6.266  .000  
2  (Constant)  5.247  .637 
 8.239  .000 
QualslpRecoded  .751  .158  .247  4.748  .000  
how stressed over last month  .664  .074  .467  8.991  .000  
a. Dependent Variable: HADS Anxiety 
B. Write one paragraph of a Results section to report the findings in APA style. Be sure to include the significant and nonsignificant findings and the statistics from each of the analyses.
A multiple regression was used to assess the ability of two variables (quality of sleep recoded and how stressed) to predict anxiety. The total variance explained by the model was 33.5%, F (2,262)= 66.01, p<.001. The greatest significant predictor for this model was stress over the last month (beta= .47, b= .66, p<.001), followed by quality of sleep recoded (beta=.25, b= .75, p<.001). For every unit increase in stress, anxiety increased by 0.66 and for every unit of increased quality in sleep, anxiety decreased by 0.75.
C. Suppose you want to explain the results of this analysis to a friend who is not in the class. Translate the results paragraph into a onesentence takehome message.
The higher level of stress a person feels the more anxiety they feel and the more quality sleep a person gets the lower the anxiety they feel. So basically sleep more and reduce stress to feel less anxiety.
D. Compare and contrast the results from Model 1 and Model 2. (a) Describe and explain why there is a difference in Rsquare across the two models. (b) Describe and explain why there are differences in the intercept and the coefficient for quality of sleep recoded across the two models.
A) Model 1 accounts for the total variance of anxiety that is only attributed to quality of sleep recoded which is 12.7 percent of the total variance of anxiety. Whereas, model 2 accounts for the total variance of anxiety that is attributed to the combined effects of both quality of sleep recoded and how stressed over the last month, which accounts for 33.5 percent of the total variance in anxiety.
B) The intercept in model 1 is 9.31 which is the predicted score on anxiety when quality of sleep recoded alone =0, the negative sign shows that as quality of sleep increases by one unit, anxiety decreases by 1.10, and Beta tells us that for every standard deviation increase in quality of sleep, anxiety decreases by .36 standard deviations. Whereas the intercept in model 2 is 5.25 which is the predicted score on anxiety when both quality of sleep recoded and how stressed over the last month combined =0, the negative sign shows that as quality of sleep increases by one unit, anxiety decreases by .75, and Beta tells us that for every standard deviation increase in quality of sleep, anxiety decreases by .25 standard deviations. The positive sign shows us that as stress increases by one unit, anxiety increases by .66, and Beta tells us that for every standard deviation increase in stress, anxiety increases by .47 standard deviations.
Darius Muniz
Professor Godfrey
Research Methods II
November 13th, 2011
Step 1:
Total Variance Explained  
Factor  Initial Eigenvalues  Extraction Sums of Squared Loadings  Rotation Sums of Squared Loadingsa  
Total  % of Variance  Cumulative %  Total  % of Variance  Cumulative %  Total  
dimension0  1  4.214  35.116  35.116  2.576  21.467  21.467  2.310 
2  2.268  18.897  54.013  2.849  23.739  45.206  3.422  
3  1.169  9.742  63.756  1.093  9.111  54.317  2.045  
4  .759  6.326  70.081 



 
5  .688  5.730  75.812 



 
6  .630  5.251  81.062 



 
7  .556  4.635  85.698 



 
8  .478  3.982  89.680 



 
9  .404  3.363  93.043 



 
10  .351  2.929  95.972 



 
11  .302  2.513  98.485 



 
12  .182  1.515  100.000 



 
Extraction Method: Maximum Likelihood.  
a. When factors are correlated, sums of squared loadings cannot be added to obtain a total variance. 
A) I have decided to retain three Factors based on the eigen values chart in which these factors are greater than 1. Factors 1, 2, and 3 made up the majority of the percentage of variance explained totaling upwards of 63 percent of the total variance and factor three was the cutoff point for where the elbow of the scree plot begins.
Step 2:
Structure Matrix  
 Factor  
1  2  3  
1. interested  .171  .663  .207 
2. upset  .485  .240  .589 
3. scared  .975  .223  .390 
6. determined  .109  .498  .008 
10. hostile  .260  .114  .716 
7. active  .278  .565  .209 
11. irritable  .362  .215  .824 
12. inspired  .156  .770  .136 
13. attentive  .183  .676  .212 
14. afraid  .833  .221  .386 
15. excited  .099  .658  .099 
17. enthsiastic  .214  .848  .214 
Extraction Method: Maximum Likelihood. Rotation Method: Oblimin with Kaiser Normalization. 
Factor Correlation Matrix  
Factor  1  2  3  
dimension0  1  1.000  .250  .425 
2  .250  1.000  .215  
3  .425  .215  1.000  
Extraction Method: Maximum Likelihood. Rotation Method: Oblimin with Kaiser Normalization. 
A) I labeled factor 1 as Fearfulness due to the high loadings of the variables scared and afraid and how these variables were representative of fearfulness. Factor 2 was labeled as Positive Motivation due to the high loadings of variables that were representative of the construct of positive motivation including interested, determined, active, inspired, attentive, excited, and enthusiastic. Finally I labeled factor 3 as Aggression due to the high loadings of hostile and irritable and how these variables were representative of aggression.
B) The variable upset double loaded onto factor 1 at .485 as well as factor 3 at .589. Though this variable had a higher loading for factor 3 (aggression) I chose not to include this variable in either factor 1 nor factor 3 due to its ability to be included in both factor 1 (fearfulness) and factor 3 (aggression). This fact made placing this variable in either fearfulness or aggression a difficult and ambiguous decision.
C) Factor 1 (Fearfulness) was negatively correlated with factor 2 (Positive Motivation) and positively correlated with factor 3 (Aggression). These correlations are as expected indicating: as a person’s fearfulness increases as well as a person’s aggression level, a person’s positive motivation level decreases. Also that a person’s fearfulness is positively correlated with a person’s aggression level indicating as fearfulness increases so does a person’s aggression level.
Step 3:
A) Seven variables (items) were used to measure participants positive motivation on a 5point Likert scale ranging from 1 (very slightly) to 5 (extremely) (α = .85).
Step 4:
Descriptive Statistics  
 N  Minimum  Maximum  Mean  Std. Deviation 
PositiveMotivation  436  1.00  5.00  3.4014  .75826 
Valid N (listwise)  436 




Login