Presentation on theme: "SPSS Session 1: Levels of Measurement and Frequency Distributions"— Presentation transcript:
1 SPSS Session 1: Levels of Measurement and Frequency Distributions
2 Learning Objectives Measurement of data Levels of measurement Measures of Central TendencyMeasures of Dispersion
3 Review from Lecture 7Identified and defined levels of measurement and measures of central tendencyDescribed situations in which different levels of measurement and measures of central tendency were useful and appropriateCalculated frequency, percentage, range, measures of central tendencyCritiqued and justified their use for the exercise problems
4 Levels of MeasurementThe level of measurement used in collecting data determines the statistical techniques which can be used in analysis.Levels of measurement:NominalOrdinalInterval/Ratio
5 Nominal Level Measurement Classifying items into groupsNo implied value of the groups as in a hierarchy or quantitative valueIn the dataset from our child protection study, nominal variables includeGender of the respondent and childmale or femaleGeneral Health Questionnaire elevated scoresSubclinical score or clinically elevated score
6 Ordinal Level Measurement Classifying values of a variable in an orderQuantitatively ordered items with an implied qualitative orderAn example is a Likert scale question with possible responses:1. Never, 2. Sometimes, 3. Occasionally, 4. Often, 5. AlwaysAn example in our child protection study of an ordinal variable:Previous_Involvement - Have Social Services been involved with this child/ family previously?1. Yes – Long standing involvement2. Yes – Occasional involvement3. No – No previous involvement
7 Interval/Ratio Level Measurement Interval/Ratio level variables have equal units between variables and offer a range of possible values in that variableAge, time, and weight are examplesExamples in our child protection study of interval/ratio variables are:GHQ-12 total scoreFES scoresWAI scoresAge of childOther total scores from standardized measures
9 Frequency Distributions A distribution provides a summary of how the data exists on a range of possible or actual scores.A frequency distribution combines all of the like values of a variable and graphical groups them.Which is to say how many times a value was recorded in a variableCharts such as a histogram provide a visual display of a frequency distribution where the frequencies of similar values in a variable are grouped
10 Frequency Distributions In our child protection study:Gender of the childAge of the childPrevious involvement with social servicesGender of the child is a nominal variablePrevious Involvement with social services is an ordinal variableAge of the child is an interval/ratio variableUse the Analyze Menu in SPSS to find “Frequencies”
12 Select the variables from the list on the left and place in the “Variable(s)” list on the right.
13 Click on “Statistics” and select “Mean”, “Median”, “Mode”, “Standard Deviation”, “Minimum”, and “Maximum”Click “Continue”
14 Click on “Charts”, and select “Histograms” with “Show normal curve on histogram” Click “Continue”
15 Frequency Distributions Click “OK” for the Frequency Distributions and the descriptive statistics for these three variables.The results will appear in a new Output window
16 In the first table, the descriptive statistics for the three variables are displayed.
17 Frequency Tables and Histograms The next three slides give the Frequency Table and Histogram for each of the three variables we selected.When comparing the tables to the histograms, look to see how similar values are combined and visually displayed in the chart.Also, compare the distribution in the histogram to the curve of that the distribution would be if the variable were normally distributed.
21 Measures of Central Tendency Mean – summing all the scores in a dataset and dividing by the total number of scores. Provides an average score.Median – The middle most score in a list of scoresMode – The most frequent or common score in a list of scores
22 Measures of Central Tendency From these results, we can see that the mean of the children in the study was 7.69 years.Remember that it is inappropriate to take means (averages) of nominal or ordinal variables, thus the Means and Std. Deviation scores for Child Gender and Previous Involvement should be ignored.