Results

Data from Ong et al. (2011). The authors investigated whether extraversion and narcissism predict the frequency of Facebook status updates and Facebook profile picture ratings.

Frequency of status updates

The first linear model looks at whether narcissism predicts, above and beyond the other variables, the frequency of status updates. To do this, drag the outcome variable status to the Dependent Variable box, then drag the variables age, grade, extraversion, and narcissism to the Covariates box, and the variable sex to the Factors box. Then, under the Model tab, define the three models as follows. In Model 0, put age, sex, and grade. In Model 1 add extraversion, and then in Model 2 add narcissism:

Model Summary - status
Model R R² Adjusted R² RMSE R² Change F Change df1 df2 p
M₀ 0.200 0.040 0.028 2.451 0.040 3.426 3 247 0.018
M₁ 0.236 0.056 0.040 2.436 0.016 4.133 1 246 0.043
M₂ 0.299 0.090 0.071 2.396 0.034 9.078 1 245 0.003
Note.  M₀ includes age, sex, grade
Note.  M₁ includes age, grade, extraversion, sex
Note.  M₂ includes age, grade, extraversion, sex, narcissism
ANOVA
Model   Sum of Squares df Mean Square F p
M₀ Regression 61.732 3 20.577 3.426 0.018
  Residual 1483.712 247 6.007  
  Total 1545.444 250  
M₁ Regression 86.250 4 21.563 3.635 0.007
  Residual 1459.194 246 5.932  
  Total 1545.444 250  
M₂ Regression 138.384 5 27.677 4.819 < .001
  Residual 1407.061 245 5.743  
  Total 1545.444 250  
Note.  M₀ includes age, sex, grade
Note.  M₁ includes age, grade, extraversion, sex
Note.  M₂ includes age, grade, extraversion, sex, narcissism
Coefficients
95% CI
Collinearity Statistics
Model   Unstandardized Standard Error Standardizedᵃ t p Lower Upper Tolerance VIF
M₀ (Intercept) 3.383 3.674 0.921 0.358 -3.852 10.619  
  age -0.033 0.309 -0.014 -0.107 0.915 -0.642 0.576 0.236 4.233
  sex (2) -0.775 0.327 -2.370 0.019 -1.420 -0.131 0.936 1.068
  grade -0.444 0.388 -0.149 -1.145 0.253 -1.208 0.320 0.229 4.365
M₁ (Intercept) 0.830 3.861 0.215 0.830 -6.775 8.434  
  age -0.006 0.308 -0.002 -0.019 0.985 -0.612 0.600 0.236 4.241
  grade -0.486 0.386 -0.163 -1.259 0.209 -1.246 0.274 0.228 4.378
  extraversion 0.052 0.025 0.127 2.033 0.043 0.002 0.101 0.977 1.024
  sex (2) -0.691 0.328 -2.110 0.036 -1.337 -0.046 0.921 1.085
M₂ (Intercept) 0.650 3.799 0.171 0.864 -6.833 8.134  
  age -0.010 0.303 -0.004 -0.033 0.974 -0.606 0.586 0.236 4.241
  grade -0.522 0.380 -0.175 -1.375 0.170 -1.271 0.226 0.228 4.382
  extraversion 0.011 0.028 0.028 0.394 0.694 -0.045 0.067 0.758 1.320
  sex (2) -0.943 0.333 -2.831 0.005 -1.599 -0.287 0.864 1.158
  narcissism 0.066 0.022 0.212 3.013 0.003 0.023 0.110 0.752 1.329
ᵃ Standardized coefficients can only be computed for continuous predictors.

So basically, Ong et al.’s prediction was supported in that after adjusting for age, grade and sex, narcissism significantly predicted the frequency of Facebook status updates over and above extroversion. The positive standardized beta value (.21) in the Coefficients table indicates a positive relationship between frequency of Facebook updates and narcissism, in that more narcissistic adolescents updated their Facebook status more frequently than their less narcissistic peers did. Compare these results to the results reported in Ong et al. (2011). The Table 2 from their paper is reproduced at the end of this task below.

Profile picture ratings

OK, now let’s fit the second model to investigate whether narcissism predicts, above and beyond the other variables, the Facebook profile picture ratings. Drag the outcome variable profile to the Dependent Variable box, then define the three Models as follows. In Model 0 put age, sex and grade, then in Model 1 add extraversion, and finally in Model 2 add narcissism.

Model Summary - profile
Model R R² Adjusted R² RMSE R² Change F Change df1 df2 p
M₀ 0.177 0.031 0.016 3.574 0.031 2.047 3 189 0.109
M₁ 0.395 0.156 0.138 3.346 0.124 27.648 1 188 < .001
M₂ 0.493 0.243 0.223 3.177 0.087 21.562 1 187 < .001
Note.  M₀ includes age, sex, grade
Note.  M₁ includes age, grade, extraversion, sex
Note.  M₂ includes age, grade, extraversion, sex, narcissism
ANOVA
Model   Sum of Squares df Mean Square F p
M₀ Regression 78.457 3 26.152 2.047 0.109
  Residual 2414.548 189 12.775  
  Total 2493.005 192  
M₁ Regression 388.023 4 97.006 8.664 < .001
  Residual 2104.982 188 11.197  
  Total 2493.005 192  
M₂ Regression 605.642 5 121.128 12.001 < .001
  Residual 1887.363 187 10.093  
  Total 2493.005 192  
Note.  M₀ includes age, sex, grade
Note.  M₁ includes age, grade, extraversion, sex
Note.  M₂ includes age, grade, extraversion, sex, narcissism
Coefficients
95% CI
Collinearity Statistics
Model   Unstandardized Standard Error Standardizedᵃ t p Lower Upper Tolerance VIF
M₀ (Intercept) 8.782 5.689 1.544 0.124 -2.439 20.003  
  age 0.150 0.475 0.043 0.317 0.752 -0.787 1.088 0.271 3.684
  sex (2) 1.290 0.550 2.348 0.020 0.206 2.375 0.981 1.020
  grade 0.099 0.603 0.023 0.163 0.870 -1.091 1.289 0.269 3.718
M₁ (Intercept) -3.461 5.812 -0.596 0.552 -14.927 8.004  
  age 0.365 0.447 0.106 0.818 0.415 -0.516 1.246 0.269 3.715
  grade -0.245 0.569 -0.056 -0.430 0.668 -1.366 0.877 0.265 3.768
  extraversion 0.224 0.043 0.356 5.258 < .001 0.140 0.307 0.980 1.021
  sex (2) 1.475 0.516 2.860 0.005 0.458 2.493 0.976 1.024
M₂ (Intercept) -3.169 5.519 -0.574 0.566 -14.056 7.718  
  age 0.337 0.424 0.097 0.794 0.428 -0.500 1.174 0.269 3.716
  grade -0.258 0.540 -0.059 -0.478 0.633 -1.323 0.807 0.265 3.768
  extraversion 0.104 0.048 0.166 2.176 0.031 0.010 0.199 0.697 1.435
  sex (2) 0.582 0.526 1.106 0.270 -0.456 1.620 0.846 1.183
  narcissism 0.173 0.037 0.366 4.643 < .001 0.099 0.246 0.653 1.531
ᵃ Standardized coefficients can only be computed for continuous predictors.

These results show that after adjusting for age, grade and sex, narcissism significantly predicted the Facebook profile picture ratings over and above extroversion. The positive beta value (.37) in the Coefficients table indicates a positive relationship between profile picture ratings and narcissism, in that more narcissistic adolescents rated their Facebook profile pictures more positively than their less narcissistic peers did. Compare these results to the results reported in Table 2 of Ong et al. (2011) below: