Get these flashcards, study & pass exams. For free! Even on iPhone/Android!

A collection of your data representing all of the information (fields or variables) collected about each case (record or observation)

Your dataset may be created based on a survey that you have created. Or you could be collecting data from participants in a study.

Or you could download a dataset from a standard data collection agency (example: IPEDS).

Or you may be given an existing dataset from a research study.

A single piece of information that you have collected.

One "field" of information.

Each variable will get a "NAME" which should be short for easy reference in analysis.

Examples:
student's age
person's gender
person's height
person's weight
percent of Asian students in a school
average teacher salary in a school

One "record" in your dataset and all of the associated fields or variables.

Can also be referred to as a "case".

Examples:
All of the information representing a single student
All of the information representing a single teacher
All of the information representing a single school

Depending on your research study, an observation may be based on an individual or an entity (school, town, state, community).

Quick reference guide to your dataset

"Translation" handbook for your variables

What will be in your CODEBOOK?

Each variable name
A description of the variable

For categorical variables:
VALUES and VALUE LABELS
COUNT and PERCENT for each VALUE

For scale variables:
MEAN and STANDARD DEVIATION

The variable that is used to uniquely identify the observation or case.

Examples:
Student ID number (when observation represents a student)
Teacher ID number (when observation represents a teacher)
School ID number (when observation represents a school)
County ID (when observation represents a county)


Never used in analysis.
Always a NOMINAL variable.

Classifier/Identifier or Categorical variable

Categorical -> Non-ordered values
Measure of belonging without quantifying
Describe characteristics

Examples:
Race
Ethnicity
Gender
Eye Color

Categorical variable -> Ordered groups
Difference between each group is not necessarily equal

Examples:
Age groups
Under 20
20-29
30-39
40-49
50-59
60 and over

Continuous variables
Difference between values is always the same and meaningful

Examples:
Age
Height
Weight

In SPSS, you want your variable values to take on numeric values.

So if you were collecting eye color, you would assign "codes" to different colors

1 = blue
2 = green
3 = brown
4 = hazel
5 = other

and then store the number value associated with the color in the dataset

The assignment above would be recorded in your CODEBOOK for reference. And assigned to the VARIABLE LABEL.

In addition to having variable values and meanings in your codebook, you want to assign variable labels within SPSS. This will allow analysis to use the words, instead of the numbers.


For example, if eye colors were coded as:
1 = blue
2 = green
3 = brown
4 = hazel
5 = other

The output could use the words, "blue", "green", "brown, "hazel" and "other", as long as you provide these LABELS.

Sometimes you don't have all of the fields or variables associated with a record or observation. Or, you may allow your respondents to not answer, or to answer with "not applicable".

You always want to code any "non-answers" and tell SPSS that these should be treated as missing data. This will make SPSS ignore those values in any analysis.

Example: Eye Color
1 = blue, 2 = green, 3 = brown, 4 = hazel, 5 = other
96 = Not applicable
99 = Missing

Both 96 and 99 should be designated as MISSING VALUES.

SPSS also uses the "DOT" to represent SYSTEM-MISSING values. These will always be treated as missing without any special designation. Leaving a field blank when entering data will treat it as missing.

Also known as the Average

How to calculate:
Take all the values, add them up and divide by the number of items.

Only calculate the MEAN on scale variables or Likert scale responses.

Sample data:
1, 3, 5, 6, 6, 8, 10, 13

52 / 8 = 6.5  is the MEAN

The middle term

How to Calculate:
Order all of the values form highest to lowest and then take the value in the middle of the group.

Should only be calculated on scale variables.

Sample data:
1, 3, 5, 6, 6, 8, 10, 13

6 is the MEDIAN

The most common value

Take each unique value and find out how many times it occurs. Take the one that occurs most frequently.

Should only be calculated on scale variables.

Sample data:
1, 3, 5, 6, 6, 8, 10, 13

6 is the MODE

A table showing each unique variable value and the number (COUNT, N) of times it occurs in your dataset  and the percent of the dataset that it represents.

Also shows the VALID PERCENT, which eliminates the MISSING VALUES.

Gender	N	Percent
Male	8	40%
Female	12	60%

A representation of the "spread" of the data.

Should always be presented with the MEAN