Here is a case when you are comparing the mean of your sample with a known mean, it is much like doing the single-sample T-Test http://p.web.umkc.edu/pruer/which_test/Single_T.html, but in this case you have data that is not parametric.

Because the 1 - 10 scale of the Perception of Safety variable has dubious normality, owing to it is not a true number like, how many year have you worked in social work, or what is your age. We should use the nonparametric test.

For the Sake of Demonstration Only we will say that the Perception of Safety scale was also asked in various other studies and the population mean had been normed on it, and that poplation mean was μ = 2.75, we just use the wilcox.test, using something like the example below.

The ‘with’ command tells R to use the SWsafety dataset, the wilcox.test tells R what to do, the ‘PerceptionSafety’ tells R which vector or variable to use, the ‘alternative’ command could be ‘two.sided’ or either greator or less, finally the ‘mu =’ argument tells R which value to test against. Remember to not forget to put all the parantheses and commas in the right place.

with(SWsafety, wilcox.test(PerceptionSafety, alternative='two.sided', mu=2.75))
## 
##  Wilcoxon signed rank test with continuity correction
## 
## data:  PerceptionSafety
## V = 1399, p-value = 0.01595
## alternative hypothesis: true location is not equal to 2.75

Since we have a significant P value of 0.01595, we would reject the H:0 and say that our sample differs from those other studies.