**Using SPSS Computerized Statistical Tool**

**Part A**

- Statistics can be best understood if researchers get to understand the basics of statistics. This question will talk about the four measurement scales and types of central tendency mean, median, or mode that can be used to best describe each level of measurement. Nominal scale is called labeling variables, and do not have quantitative value. This level of measurement have no overlap with no numeric significance. Examples include gender, race, and place of residence among others. This level of measurement can be best described by mode, which will give researchers the chance to identify the most appearing name or label (Trochim 12).

Ordinal scale is also known as rank sale where variables are placed in a certain orderwhile the difference between the variables is not known. Examples of its application are in Likert scale, such as measuring satisfaction, happiness, and discomfort among others. It can be best described by median.

Interval level of measurement is numerical where the order is distinct and there is a difference between the values. Examples include time and temperature in Celsius. However, this level of measurement does have true zero values. Both mode, median and mean can be used to describe the data (Trochim 26).

Lastly, ratio scale or level of measurement is characterized by order, exact difference of value between units and having an absolute zero value. Central tendency that can be applied in describing ratio data include median, mode and mean.

- A measure of dispersion isaimed at measuring how spread out a set of data is. For ratio scale, measures of dispersion that can be used to describe it include standard deviation, interquartile range, range and coefficient of variation.

When dealing with the interval level of measurement, the best measure of dispersion that can best describe it include standard deviation, interquartile range and range. There is no known measure of dispersion that can be used to describe nominal and interval level of measurement (Trochim 45).

**Part B**

Topic: An investigation into the relationship between absenteeism and job related attitudes.

**Hypotheses**

- There is a link between absenteeism and the lack of pro work-life policy.
- Motivated employees take less sick-leave.
- Companies regulated by collective agreement record less absenteeism.
- HRM can meet employee motivation, expectations, and, as a result reducing absenteeism.
- There is a relationship between high staff turnover rate and absenteeism.

This hypothesis was tested using correlation analysis, generation of a scatter-plot. From, the table below, it is apparent that there is a weak negative significant relationship between age and the idea or thought that it is better for men to work and for women to tend homes, r=-0.155, p=0.000 (Table 18). This suggests that the older one is, the stronger is the thought or the idea that it is better for men to work and for women to tend homes.

Table 18 Correlations |
|||

Age of Respondent | Better for Man to Work, Woman Tend Home | ||

Age of Respondent | Pearson Correlation | 1 | -.155^{**} |

Sig. (2-tailed) | .000 | ||

N | 2013 | 1300 | |

Better for Man to Work, Woman Tend Home | Pearson Correlation | -.155^{**} |
1 |

Sig. (2-tailed) | .000 | ||

N | 1300 | 1308 | |

**. Correlation is significant at the 0.01 level (2-tailed). |

*Work Cited:*

*Work Cited:*

- Trochim, William.
*Descriptive statistics: Research Methods Knowledge Base*.Oxford University: Oxford University Press, 2003. Print.