This guide is intended to support the data analysis work that is an integral part of the Measurement and Evaluation Course. It is essential to acquire a firm grasp of the basics (descriptive statistics) since they will be used throughout the course for a wide array of analytical purposes.
Quick Links to sections of this guide:
The information presented in each section provides both context (when to use) and menu paths within SPSS to follow to execute various analyses.
"Through and through the world is infested with quantity: To talk sense
is to talk quantities. It is no use saying the nation is large - how
large?
It is no use saying that radium is scarce - how scarce? You cannot evade
quantity.
You may fly to poetry and music, and quantity and number will face you
in your rhythms and your octaves."
When you first open SPSS you will notice that on the bottom of the screen are two tabs. One is the data view the other is the variable view.
Data View. From the data view you enter your data. Each horizontal line is for data pertaining to an individual. Each vertical column is for data pertaining to a variable.
Variable View: From the variable view, you provide information pertaining to each variable in your data set. This will include providing:
Variable name: Should be a short descriptive name - no spaces permitted
Variable label: A longer descriptive phrase to describe the variable.
Value labels: For categorical data (e.g. gender) where the numbers represent categories, the values column is where you specify which category each number represents. For example Male = 0; Female = 1.
Level of Measurement: Categorical (data just represents categories); Ordinal (data represents categories in a meaningful order); Scale (refers to interval and ratio scaled data) - generally meaning you are working data that represent actual scores rather than just categories.
Adding Case/ID numbers
When you create a new data file it should include a variable named
ID. IDs can be numbers or names (e.g., cereals, pet name in hula
file). To have SPSS create the ID numbers:
You'll be able to see your new ID variable in SPSS's Data View.
To see a summary of the information in a data file displayed in the output area of SPSS:
Notice that the information produced in the output file is essentially the same as that in the variable view. The information will be displayed in two parts: the Variable Information and the Variable Values.
Following entry of data into the SPSS spreadsheet it is
important to check for errors. For example, consider the variable GENDER
with value labels of 1 for male and 2 for female. It is reasonable to
assume that a typing error could result in entries of other than a 1 or
2. One way to detect this error is to have SPSS produce a frequency
distribution table for this variable. It might look like this:
Gender | frequency | ||
Male | 35 | ||
Female | 41 | ||
3 | 6 | ||
6 | 2 |
This table makes it clear that 8 of the entries are erroneous. For six subjects the value 3 was entered for gender and for another two subjects the value 6 was entered. With the errors detected, you have two options:
1st, get a frequency distribution table for all variables and all cases in the data file:
2nd, if errors detected that are clearly values outside what is acceptable for a variable:
Note: If you encounter a situation where a value is
inappropriate but only for a particular person in the data set you will
not be able to use the 'missing values' feature in the variable view
section of SPSS. Instead you will need to find the incorrect value(s) in
the data view and delete them manually from the data file. For example,
consider the situation where you have obtained two heart rates. One
resting and the other one minute after jogging in place. If for one of
the cases the two values were 128 and 128 that seems likely to be an
error since the resting heart rate is quite high and the exercise heart
rate is unlikely to be the same as the resting heart rate. If you don't
have access to the original data so you can re-enter the correct values
then you need to delete these values from the data file. But since 128
may be a legitimate value for some other cases you can't just assign it
as a missing value from the variable view. You need to go into the data
view, find this case and delete each 128 leaving blank cells for for
this particular person/case.
Note: Constructing frequency distribution tables
for every variable for the purpose of error checking is
important to complete prior to initiating any analytical work.
|
Regardless of the nature of the variable, it is often useful to condense information before reporting it. For example: Assume you collected information on years of education in 5 categories (< High School, High School, some college, Bachelor’s degree, > Master’s degree) but only wanted to report the proportion of people with no college work and those with at least some college work. You would not want to manipulate the original variable so you would first create a new variable then recode the new variable.
In situations where you have component information and you need for example a total for each individual, a new variable needs to be created. This is easily done using the compute feature in SPSS.
There are times when it's useful to transform raw scores to standardized scores with a fixed mean and standard deviation. SPSS can do the tranformation from raw scores to Z scores (which have a mean of 0 and Standard deviation of 1).
Z scores are a type of standardized score. Their particular feature is that they have a mean of zero and standard deviation of one. Standard scores tell you how many standard deviation units above or below the mean a value falls.
There are times when you need to conduct an analysis on a portion of a data set (e.g., just the women) rather than the whole group. When this is the case, you first select the group you want to conduct the analysis on and then proceed to do the analysis (e.g., central tendency) you need for just that group. REMEMBER: when done, undo the selection so all cases are available for subsequent analyses.
Frequently due to the nature of the group that measures have been obtained from, analyses on a subset of the entire group are of interest. When this is the case you first identify the subset (select cases) then proceed with the analysis.
In situations where you would like to conduct the same analysis (e.g., correlation, reliability) on subsets of a group (e.g., males and females) you could use the split file feature in SPSS. When analyses need to be repeated on all groups that make up a variable (e.g. gender: males/females) the split file feature is ideal to use. For example you may want to look at the correlation between exercise frequency and cholesterol level for men then for women. You could of course use the 'select cases' procedure above first for the males then repeat for females. However, the split file feature lets you do the two analyses at the same time.
Summarizing group information is typically the first step in the search for patterns, highlights, and meaning in a data set. Summary information can be presented both visually with the use of graphs and in the form of summary statistics. This section will focus on:
The connection between the level of measurement for data and the selection of appropriate statistics to summarize that data is an important one. The table below provides some guidance on what statistics are approriate for each level of measurement.
Level of Measurement | Applicable Statistics |
Nominal/Categorical | Percentages, Mode |
Ordinal | Percentages, Mode, Median* |
Interval | Mean, Median, Mode, Standard Deviation, Range |
Ratio | Mean, Median, Mode, Standard Deviation, Range |
For categorical and ordinal data the construction of frequency distribution tables is an excellent way to summarize group information.
If you were to make a frequency distribution table by hand you would simply list each category/value observed followed by a count (also called absolute frequency) of the number of individuals in that category. An additional column called the relative frequency is often useful since it notes the percentage of the group in a particular category. For example:
Gender | f | rf | ||
Male | 28 | 48% | ||
Female | 30 | 52% |
f: absolute frequency - count
rf: relative frequency - count/N (100) - record as %
To get a frequency distribution table for all cases in the data file:
To get a frequency distribution table for a subset of cases in the data file:
With subgroup now selected:
Remember to go back through data menu to reselect all cases before starting analyses where all cases are needed.
Note: You would not construct frequency distribution tables for continuous data when the intent is to summarize information. The reason is that such data can take on a great number of values and since each value is listed in a frequency distribution table little summary may be accomplished. Measures of Central Tendency and Variability are much more useful in summarizing group information for interval and ratio scaled data. |
For categorical and ordinal data the construction of crosstabulation tables is an excellent way to cross-reference summary information for two or more variables.
If you were to make a crosstabulation table by hand you would in rows list each category/value of one variable and in columns list each category/value of a second variable. The table then would contain a count of the number of individuals in cells representing the various combinations of values for the two variables. For example, you might want to combine in one table gender (categorical) and age group (ordinal).
Age Group | ||||
20-25 | 26-30 | 31-35 | ||
Male | 28 | 20 | 15 | |
Gender | ||||
Female | 30 | 18 | 20 |
From this table you can see that 28 of the subjects were male and in the youngest age group, and 18 of the subjects were female and in the middle age group.
Step Summary to break down by a 3rd variable (layered cross tabulation)
Note: You would not construct crosstabulation tables for continuous data when the intent is to summarize information. The reason is that such data can take on a great number of values and each value would be listed in a crosstabulation table. Therefore little summary may be accomplished. Measures of Central Tendency and Variability are much more useful in summarizing group information for continuous variables. |
Measures of central tendency summarize data by identifying where the center of a distribution of scores is. Measures of variability summarize data by quantifying the spread or dispersion of scores around the center.
For categorical and ordinal data with few categories, the Mode (though not an optimal measure) is an acceptable measure of central tendency however, discrete data is best summarized with a frequency distribution table.
For data at least interval scaled, the Median and Mean are appropriate measures of central tendency. If the distribution of scores is skewed the Median is the best measure of central tendency. The most common measure of variability is the standard deviation and is appropriate for use with data at least interval scaled.
In addition to being used to summarize a data set, measures of central tendency and variability are critical components of other statistical procedures.
REMEMBER, you must check the shape (obtain histogram under graphs option) of the distribution of scores to decide what measure of central tendency is appropriate. If the shape is clearly skewed then you need to obtain a median.
Central Tendency & Variability for data from one variable:
To get measures of central tendency and variability for data from one interval/ratio scaled variable broken down by one discrete variable:
To break the analysis down by a 2nd categorical variable (layered compare means):
Useful for conveying relative information about an individual are percentiles (raw score with specified percentage below it). Specific percentiles can be requested under the statistics option under frequency distribution tables.
There are several types of correlation coefficients to choose from. The choice is based on the nature of the data being correlated.
Pearson Product Moment Correlation | Use when both variables have interval or ratio scaled data |
Phi | Use when both variables are discrete and data are dichotomous |
Cramer's V | Use when both variables are discrete and at least one of the variables has more than two categories |
Kendall's Tau | Use when both variables have ordinal data |
Point Biserial Correlation | Use when one variable has interval or ratio scaled data and the other a true dichotomy |
The PPMC can be used to describe the strength and direction of the linear relationship between two continuous variables. When two variables are not linearly related, the PPMC is likely to underestimate the true strength of the relationship. A graph of the x and y values can show whether or not the relationship is linear.
Kendall's Tau can be used to describe the strength and direction of the relationship between two ordinal variables. It is a rank-order correlation coefficient (as is PPMC) and can convey the extent to which pairs of values (x,y) are in the same rank order.
Phi can be used to describe the strength of the relationship between two variables each with data that is dichotomous.
Cramer's V can be used to describe the strength of the relationship between two discrete variables.
Phi and Cramer's V are signed numbers between -1 and 1 where zero represents no relationship.
The Point Biserial Correlation can be used to describe the strength of the relationship between one continuous (interval or ratio scaled) variable and one dichotomous variable. The point biserial correlation coefficient is useful in detecting a pattern in group measures (e.g., one group's scores tending to be higher than another group). The sign carries little meaning. It only indicates which group tended to have higher scores. The point biserial coefficient is a signed number between -1 and 1 where zero represents no relationship.
The computational formula for the point biserial coefficient is
Where:
X0 = mean of x values for those in category 0
X1 = mean of the x values for those in category 1
Sx = standard deviation of all x values
P0 = proportion of people in category 0
P1 = proportion of people in category 1
To obtain the components you need from SPSS so you can do Point Biserial by hand, you would use the compare means feature in SPSS:
Graphs are the visual counterparts to descriptive statistics and are very powerful mechanisms for revealing patterns in a data set. In addition, when used appropriately in a report they can highlight trends and summarize pertinent information in a way no amount of text could.
When summarizing categorical data, pie or bar charts are the most efficient and easy to interpret though line graphs may be more helpful particularly at times when trying to draw attention to trends in the data. For continuous (interval or ratio scaled) data, histograms are a good choice, easily constructed and simple to interpret. When attempting to represent visually the relationship between two continuous variables a scattergram can be used.
To create simple bar for categorical and ordinal (with few categories) data:
To create a scattergram (two continuous variables)
To create a histogram (interval or ratio scaled data):
To create histograms (interval or ratio scaled data) for separate groups from a discrete variable:
To break down by a 2nd discrete variable:
Following administration of an exam comprised of multiple choice items, statistical examination of the quality of the items with respect to difficulty and ability to distinguish among ability levels can be done.
Of interest is what proportion of the group got the item correct. While SPSS does not provide this information directly, provided you have labeled correct a one and incorrect as zero the proportion can be easily obtained.
Depending on the type and purpose of a test, evidence of criterion-related validity (e.g.,concurrent, predictive) can be obtained using a correlation coefficient.
This is examined when you are interested in the extent to which a particular measure is as good as an already established criterion measure already known to provide valid and reliable data. You determine this by correlating your scoress (x is continuous) with scores or classifications from a criterion measure (y).
The process would entail:
This is examined when you are interested in the extent to which a particular measure is a good predictor of another variable. You determine this by correlating your scoress (x is continuous) with scores or classifications from the measure you are trying to predict (y).
Depending on the type and purpose of a test, evidence of criterion-related validity of classifications (e.g., master-nonmaster) can be obtained from a correlation coefficient.
The concurrent validity of classifications is examined when you are interested in the extent to which classifications (master/non master) are correct. You determine this by correlating your classifications (x) with classifications or scores from a criterion measure (y).
This is examined when you are interested in the extent to which classifications are good predictors of another set of classifications or scores. You determine this by correlating your classifications (x) with classifications or scores from a variable you are trying to predict (y).
The primary concern here is the accuracy of measures. Reducing sources of measurement error is the key to enhancing the reliability of the data.
Reliability is typically assessed in one of two ways:
To estimate reliability you need 2 or more scores (or classifications) per person.
Note: When interpreting Cronbach's alpha or the intraclass R, a value > .70 reflects good reliability. |
If multiple cognitive and motor skills/physiological measures collected at one time or over time, you can use an intraclass coefficient to estimate reliability.
Once you have 2 scores per person the question is how consistent overall were the scores?
NOTE: In many situations reliability has been estimated incorrectly using the Pearson correlation coefficient. This is not appropriate since (1) the PPMC is meant to show the relationship between two different variables - not two measures of the same variable, and (2) the PPMC is not sensitive to fluctuations in test scores. The PPMC is an interclass coefficient; what is needed is an intraclass coefficient.
The most commonly used and appropriate reliability coefficients are the intraclass R calculated from values in an analysis of variance table and Cronbach's alpha.
Spearman Brown Prophecy Formula - when test length manipulated
There are situations where you might want to understand how changes in test length may affect reliability. When this is the case, you 1st obtain Cronbach's alpha for the 'original' length test then apply the Spearman Brown Prophecy forumula.
m = amount you need to boost/diminish test length.
R = reliability coefficient (e.g. Cronbach's Alpha)
Note: this can be particularly useful when you administer a test only once and multiple measures are not available. In this case for example with a cognitive test, the most common way of getting 2 scores per person is to split the measures in half - usually by odd/even itmes or first half/second half by time or trials for motor skills tests.
You now have the reliability of scores for the 1/2 length test. To get reliability for the full length test, use the spearman brown prophecy formula:
In this instance you are interested in the consistency of classifications from a mastery test. The two statistics of interest are the proportion of agreement (compute by hand from values in a crosstabulation table) and Kappa.
In motor skill performance settings it is often necessary to collect measures through observation. To examine the objectivity of these measures you look at the consistency of measures across observers (inter-rater consistency). Note: you may also video tape a group and have one person record measures on two occasions (intra-rater consistency).
To assess objectivity, your task, since the measures come from observations, is to examine the objectivity of the measures produced by observers (likely using a rating scale). To do this, have two people observe one group of examinees and evaluate their performance using a rating scale. The measures from the two observers (you could also videotape the group and have one person evaluate the group twice) give you two scores per person to use in the Cronbach's alpha or intraclass R formulas. The Spearman-Brown formula is not needed in this situation since test length is not manipulated.
Note: When interpreting Cronbach's alpha or the intraclass R, a value > .70 reflects good objectivity. |
In this instance you are interested in the consistency of classifications from two observers (or one observer scoring video twice). The two statistics of interest are the proportion of agreement (compute by hand from values in a crosstabulation table) and Kappa.
The data you work with can either be scores that are converted to classifications based on a cut score or direct classifications from observers. To assess objectivity, your task, since the classifications come from observations, is to examine the objectivity of the classifications produced by observers using a rating scale or checklist. To do this, have two people observe one group of examinees and evaluate their performance using a rating scale or checklist. The classifications from the two observers (you could also videotape the group and have one person evaluate the group twice) give you two classifications per person to use in the proportion of agreement and Kappa statistics.