In this glossary you will find only a very short description of each concept. If you need a more detailed explanation, please follow one of the external links.
Some of you may remember that a line on a plane can be expressed as an equation. Assume first that there are two variables, one that is measured along the plane’s vertical axis and whose values are symbolised by the letter y, and one that is measured along the horizontal axis and whose values are symbolised by the letter x. The function that represents a linear association between these two variable values can be expressed as follows: y = a + b∙x
Here, a and b symbolise constants (fixed numbers). The number b indicates how much the variable value y increases or decreases as x changes. When x increases by 1 unit, y increases by b units. To see this, assume, for instance, that x has the initial value 5. Insert this value into the equation. You get: y = a + b∙5. Then, let the value x increase by one unit to x = 5 + 1 and insert this in the equation instead of the former value. Now the equation can be written y = a + b∙5 + b∙1. Thus, by letting x increase by 1 unit, we have made y increase by b∙1 = b units. This implies that if b is equal to, say, 2, y will increase by 2 units whenever x increases by 1 unit, or if b is negative and equal to, say, -0.5, y will decrease by 0.5 units whenever x increases by 1 unit.
The latter case is illustrated in Figure A, where we have drawn a line in accordance with the function y = 7 - 0.5∙x (where we have inserted the randomly chosen value 7 for a).
As explained above, x and y are variables, which means that they can take on a whole range of different values, while 7 and -0.5 are constants, i.e. values that determine the position of the line on the plane and which cannot change without causing the position of the line to change. In figure A, the x-values range between 0 and 10, while the y-values range between 2 and 7. It is to be hoped that you realise that for each value x takes on between 0 and 10, the equation and the corresponding line assigns a unique numerical value to y as shown in the figure. Thus, if x = 0, y takes the value of 7, which is the value we have given to the constant a. Check this out by inserting 0 in place of x in the equation y = 7 - 0.5∙x and then compute the value of the expression on the right-hand side of the equals sign. (You should get y = 7.) This illustrates the important point that the constant a in the equation y = a + b∙x is identical to the value that y takes on when x = 0. From a graphical point of view (see Figure A), a can be interpreted as the distance between two points on the vertical axis, namely the distance between its zero point (y = 0) and the point where this axis and the line given by the equation meet each other. (Assuming that the vertical axis crosses the horizontal axis in the latter’s zero point.) Thus, the constant a is often called the intercept.
Now for the graphical interpretation of the constant b: We repeat the exercise from the numerical example presented above by starting from the point on the line in Figure A where x = 5 (as marked by vertical line k). We then increase the x-value by 1 unit to x = 6 as we move downwards along the line (the new x-value is marked by vertical line l). This change in x makes the corresponding value of y decrease from 4.5 (marked by horizontal line m) to 4 (marked by horizontal line n), i.e. it ‘increases’ by -0.5 units (decreases by 0.5 units).Thus, Figure A confirms what we just saw in our numerical example: b is the change in y that takes place when x increases by 1 unit, provided that the changes occur along the line that is determined by the function y = a + b∙x. If b is positive, y increases whenever x increases, and if b has a negative numerical value, y decreases when x increases.
Note also that b can be interpreted as a measure of the steepness of the line. The more y changes when x is increased by 1 unit, the steeper the line gets.
Use these relational logical operators in If statements in SPSS commands:
EQ or = | Equal to |
NE or ~= or = or <> | Not equal to |
LT or < | Less than |
LE or <= | Less than or equal to |
GT or > | Greater than |
GE or >= | Greater than or equal to |
Two or more relations can be logically joined using the logical operators AND and OR. Logical operators combine relations according to the following rules:
AND | Both relations must be true for the complex expression to be true. |
OR | If either relation is true, the complex expression is true. |
The following table lists the outcomes for AND and OR combinations.
Expression | Outcome | Expression | Outcome |
---|---|---|---|
true AND true | = true | true OR true | = true |
true AND false | = false | true OR false | = true |
false AND false | = false | false OR false | = true |
true AND missing | = missing | true OR missing | = true |
missing AND missing | = missing | missing OR missing | = missing |
false AND missing | = false | false OR missing | = missing |
SPSS acknowledges two types of missing values: System-missing and User-missing. If a case has not been (or cannot automatically be) assigned a value on a variable, that case’s value on that variable is automatically set to ‘System missing’ and will appear as a . (dot) in the data matrix. Cases with System-missing values on a variable are not used in computations which include that variable.
If a case has been assigned a value code on a variable, the user may define that code as User-missing. By default, User-missing values are treated in the same way as System-missing values.
In the ESS dataset, refusals to answer and ‘don’t know’ answers etc. have been preset as User-missing to prevent you from making unwarranted use of them in numeric calculations. If you need to use these values to create dummy variables or for other purposes, you must first redefine them as non-missing. One way to achieve this is to open the ‘Variable View’ in the data editor, find the row of the variable whose missing values you want to redefine, go right to the ‘Missing’ column, click the right-hand side of the cell, and tick ‘No missing values’ in the dialogue box that pops up. You can also use the MISSING VALUES syntax command (see SPSS’s help function for instructions). Cases with System-missing values can be assigned valid values using the ‘Recode into different variables’ feature in the ‘Transform’ menu. Be careful when you use this option, that you do not overwrite value assignments that you would have preferred to keep as they are.
Moreover, if you need to define more values as User-missing, you can use the syntax command MISSING VALUES or the relevant variable’s cell in the ‘Missing’ column in the ‘Variable View’.
An SPSS syntax is a text command or a combination of text commands used to instruct SPSS to perform operations or calculations on a data set. Such text commands are written and stored in syntax files, which are characterised by the extension .spx.
In order to run the syntax commands that have been provided with this course pack, we suggest that you first open an SPSS syntax file. Either create a new syntax file (click ‘New’ and ‘Syntax’ on the SPSS menu bar’s ‘File’ menu) or open an old one that already contains commands that you want to combine with the new commands (click ‘Open File’ and ‘Syntax’ in the ‘File’ menu, and select the appropriate file). Then find, select and copy the relevant syntax from this course pack’s website and paste it into the open syntax file window. While doing exercises, you may have to make partial changes to the commands by editing the text. Run commands from syntax files by selecting them with the cursor (or the shift/arrow key combination) before you click the blue arrow on the syntax window’s tool bar.
If you use the menu system, you can create syntaxes by clicking ‘Paste’ instead of ‘OK’ before exiting the dialogue boxes. This causes SPSS to write the commands you have prepared to a new or to an open syntax file without executing them. Use this option to store your commands in a file so that you can run them again without having to click your way through a series of menus each time. New commands can be created from old ones by copying old syntaxes and editing the copies. This saves time.
A case's (a person's) value on a variable must be given a code in order for it to be recognised by SPSS. Codes can be of various types, for instance dates, numbers or strings of letters. A variable's codes must consist of numbers if you want to use it in mathematical computations. Value codes may have explanatory labels that tell us what the codes stand for. One way to access these explanations is to open the data file and keep the ‘Variable view’ of the SPSS 'Data editor' window open. (You can toggle between 'Variable view' and 'Data view' by clicking the buttons in the lower part of the window.) In the 'Data view', each variable has its own row. Find the cell where the row of the variable you are interested in meets the column called 'Values'. Click the right end of the cell. A dialogue box that displays codes and corresponding explanatory labels appears. These dialogue boxes can be used to assign labels to the codes of variables that you have created yourself (recommended). Codes of continuously varying variables do not have explanatory labels. The meaning of the codes of such variables must be stated in the variable label.
The value labels can also be accessed from the 'Variables' option in the 'Utilities' menu or from the variable lists that appear in many dialogue boxes. Right click the variable label and click 'Variable information'.