ABSTRACT
In Chapter 2 some key principles underpinning inferential statistics were introduced. When these ideas were explained, it was at times necessary to simply accept that certain steps can be taken. How these steps can be taken was not discussed. For example, it was stated that it is possible to use sample data to calculate confidence intervals which specify, with a particular degree of certainty (e.g. 95 per cent), the range of values within which a population mean (for continuous data) and a population proportion (for categorical data) will fall. So, we might measure the well-being of 50 Californian hospital nurses and calculate that the 95 per cent confidence interval for the mean well-being of all Californian hospital nurses on a 5-point scale is between 3.2 and 3.8. We don’t know for sure what the mean wellbeing of Californian nurses is, but based on our sample data we can be 95 per cent confident that this mean lies somewhere between 3.2 and 3.8 on our measurement scale. Chapter 2 also explained that statistical tests can be used to test research hypotheses such as ‘there is an association between how much people earn and their well-being’, and ‘using an appraisal system increases job performance’. However, again no account was given as to how this can be done.
