RWeek 5: Data input in SPSS20 Input these information in SPSS 20 Employee gender age Training received How satisfied with training? 1 Male 16-19 Less than one day neutral 2 Female 30-39 More than one day but less than 2 days dissatisfied 3 Female 50-59 Between 5- 10 days Very satisfied 4 Male Over 65 More than 10 days satisficed 5 Male 20-29 More than 2 days but less than 5 days Very satisfied 6 Female 16-19 Between 5- 10 days dissatisfied 7 Male 40-49 Between 5- 10 days Very satisfied 8 Female 30-39 More than 10 days neutral 9 Female 50-59 Less than one day satisfied 10 male 16-19 More than 2 days but less than 5 days Very dissatisfied Open SPSS Step 1: Start all programs IBM SPSS statistics 20 Step 2: Click on" Type in data", then "OK" Step 3: you will see two windows: "Data View" and "Variable View" in the left corner of the screen. The first window shows digit; the second window is how to digitalize information, define variables. Step 4: click on "variable view", you can see Name: name of the variable. Type: type of variables, numeric, currency, date, or integer Width: length of the value Decimals: keep it 0 for numeric variables Label: definition of the variable such as how satisfied are you with the training? Values: value of variable: 1 very dissatisfied; 5 very satisfied Missing: how to deal with missing data, we normally give a large, negative value to warn us the missing value, such as -999 Columns: leave it 8 Align: normally choose center Measure: Scale (profit, salary etc.) Ordinal (very satisfied, satisfied etc.) Nominal (Telford, Coventry, Shrewsbury) Coding variables Step1: variable view Line 1: ID- numeric-8-0-Employee ID-none-Missing value (-999—980) (see the picture below show how to set missing value), click ok. Choose: 8-center-scale.   Line 2: look at the task table, input: Gender-numeric-8-0-gender of employee. In term of value, click cell, you see a new window open; input 1 for "value"; "label": male; click "add". Then 0 for "value" ; label "female", click "add". Click "OK" (see the picture below) Carry on to input missing, columns, align, measure (scale) This is how we digitalize information. When there are two categories of one variable, this is the way to create a dummy variable. Such as employed and unemployed; home company and international company… Line 3: There are two ways to input this data, I show you the numeric way since we want to see as many digits as possible to do statistics. For interval data, such as age group or income group, we choose a sensible mid-point to represent that group. Such as 16-19, we use 17.5 or 17 to present this group. It is the open end group which a little bit trick, you have to refer to relevant information to give it a sensible data, such as "over 65", you can input 68 for UK data since there is no default retirement age in the UK since 2014; but state pension age can be between 61 and 68 depends on year of birth. You need to write this done in the "values"; click on the value cell, a new window appears. "Value" 17; "label" 16-19, click add "Value" 25; "label" 20-29, click add "Value" 35; "label" 30-39, click add "Value" 45; "label" 40-49, click add "Value" 55; "label" 50-59, click add "Value" 68; "label" over 65, click add Here it is what it look http://wlv.ac.uk/its/downloads/Downloads.aspx Line 4. This variable can be dealt like age. Less than one day can be interpreted as 0.5; how about "more than 10 days"? You do the same for "value". Click value cell "Value" 0.5; label "less than one day" "Value" 1.5; label "more than one day but less than two days" "Value" 3.5; label "more than two days but less than five days" "Value" 7.5; label "more than five but less than 10 days" "Value" 12; label "more than 10 days" Since we have 0.5 days, we will keep one decimal for this variable Note: only for this variable, change its "decimals" as 2. This is what it looks like Line 5: This is a typical ordinal variable. It is similar to dummy variable input. In the label cell, input "satisfaction with training"; click the values cell, you will see a new window, "value" input 1; "label" input "very dissatisfied", click "add", carry on with 2, dissatisfied; 3 neutral; 4 satisfied; 5 very satisfied. Then click "ok" Data inputs Click on "Data View", look at your task table, here you transfer those information into digits via your code. Take first employee for example. ID-1; Male-1; age (16-19)-17; training (less than 1 day)-0.5; satisfaction with training (neutral)-3 This is what it looks like. The carry on with the rest Here is half way data input look like This is how it looks at the end Some remarks In this dataset, we don't need ID. But it is important to know how to input it, when you deal with more complicated dataset, such as 10 students from each school, you will need to give each school and each student an ID. Save as in your account, so you can find it next time. We also can code age group as normal variable; can you find it out how to do it? How do we know our data input is correct? For such a micro sized dataset, you can check one by one, but how about when you have several hundred observations, each observation with hundred variables? We can use certain techniques to help us to find such error, which is via data summary and descriptive analysis. We will look at it next session. However, there are so many ways for you learn; please visit Youtube, type in "descriptive analysis in SPSS". Watch at least three video clips.