Chapter 14 1. What level(s) of measurement are used for correlation (nominal, ordinal, or scale)? 2. What would be considered a strong reading for a correlation (.10 or .40)? 3. If two variables are strongly associated, does that necessarily mean that they are causally related (yes or no)? 4. True or false: The closer the correlation reading is to 1.00 or -1.00, the stronger the relationship is between two variables. 5. What is the difference between correlation and regression (that is, which one adds the power of prediction)? 6. What type of graph is used for regression?

7. Please read the analysis below in order to see how one would prepare a written analysis specific to a correlation (no additional work is required here, please only read the paragraph and review the data). When performing a correlation between the two variables respondent's education and mother's education, data findings indicated that there was a strong association between these two variables. More specifically, Pearson's correlation reading was .347. This analysis indicated a positive relationship indicating that as a mother's education increases, the the educational level of respondents also increases.

\fModel Summary Adjusted R Std. Error of Model R R Square Square the Estimate 201 .041 .040 2.728 a. Predictors: (Constant), Highest year of school completed Coefficients Standardized Unstandardized Coefficients Coefficients Model B Sid. Error Beta Sig. (Constant) 5.820 448 12.994 .000 Highest year of school -206 032 -.201 -6.448 .000 completed a. Dependent Variable: Hours per day watching TVCONBUS Confidence in major companies * CLASS Subjective class Identification Crosstabulation CLASS Subjective class identification 1 LOWER 2 WORKING 3 MIDDLE 4 UPPER CLASS CLASS CLASS CLASS Total CONBUS Confidence in 1 A GREAT DEAL Count 18 139 168 15 340 major companies % within CLASS 8.9% 15.5% 22.1% 27.8% 17.8% Subjective class identification 2 ONLY SOME Count 125 578 499 35 1237 % within CLASS 61 9% 64 5% 65.6% 64.8% 64.7% Subjective class identification 3 HARDLY ANY Count 59 179 94 4 336 % within CLASS 29 2% 20 0% 12.4% 7.4% 17.6% Subjective class identification Total count 202 896 761 54 1913 % within CLASS 100 0% 100 0% 100.0% 100.0% 100.09 Subjective class identificationChi-Square Tests Asymptotic Significance Value df (2-sided) Pearson Chi-Square 56.318 .000 Likelihood Ratio 56.956 000 Linear-by-Linear 54.955 000 Association N of Valid Cases 1913 a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 9.48.Group Statistics Std. Error Respondents sex N Mean Std. Deviation Mean R's income in constant $ MALE 798 28666.04 33344.735 1180.391 FEMALE 834 19091.00 24177.428 837.196 Independent Samples Test Levene's Test for Equality of variances t-test for Equality of Means 95% Confidence Interval of the 3td. Error Difference F Sig Sig. (2-tailed) Difference Difference Lower Upper R's income in constant $ Equal variances 34.396 .000 6.662 1630 .000 9575.039 1437.247 6755.994 12394.084 assumed Equal variances not 6.617 1419.567 000 9575.039 1447.142 6736.323 12413.755 assumedDescriptives Hours per day watching TV 95% Confidence Interval for Mean N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum LOWER CLASS 99 4.22 4.232 425 3.38 5.07 0 24 WORKING CLASS 428 2.78 2.520 122 2.54 3.02 0 24 MIDDLE CLASS 416 2.94 2:568 126 2.69 3.18 O 24 UPPER CLASS 35 2.66 2.461 416 1.81 3.50 0 10 Total 978 2.99 2.786 089 2.81 3.16 24 This column provides the average amount of television This column offers the total number that each class level watched on a daily basis. of respondents in each class level.This significance reading is less than or equal to .05 thus there are statistically significant differences with respect to television viewing by class level. ANOVA Hours per day watching TV Sum of Squares df Mean Square F Sig. Between Groups 174.149 3 58.050 7.632 000 Within Groups 7408.727 974 7.606 Total 7582.876 977