IDENTIFYING RELATIONSHPS OF RAW DATA
A. Look at the data, notice that there is a reactive relationship
between education and occupational status (high education = high
occupational status; low education = low occupational status).
B. Make a table using arbitrary categories.
1. Put highest variable for each category in upper left hand corner.
2. Determine type of relationship by diagonal which has the concentration
of the largest numbers.
TYPES OF RELATIONSHIPS
A. Direct relationships
1. Positive: These relationships (see above example) change in the same direction. As one variable increases in value, so does the other.
2. Negative: These relationships (see above example) change inversely.
If one variable increases in value, the other decreases.
B. Intervening relationships
1. Spurious: Often two variables will show a relationship, but the relationship is not between the two variables. It is a byproduct of both variables' relationships to a third intervening variable. A spurious relationship is indicated if the original strength of the measured relationship between the two variables is substantially reduced when controlling for the third. Example: Storks per square mile and Scandinavian Birth Rates. The intervening variable is Density of Population. Both storks and high birth rates are found in rural areas.
2. Suppression: As in spurious relationships, a third variable is affecting the strength of relationship. In this case, the third variable reduces the relationship measured. Example: Does racial discrimination affect the income of Japanese Americans? On first glance, Japanese Americans show no relationship; therefore, no discrimination. However, when controlling for education, a relationship emerges showing that racial discrimination still affects income levels. Because Asian Americans have higher average educational levels, education's influence suppresses the relationship of race on income levels.
3. Interaction: In this relationship, the third variable's categories affect the relationship of the original two variables differently. Example: Career Aspirations of Sociology Graduate Students by Sex with the controlling variable of GPAs. High GPA's show little differences between the sexes, but average GPA's influence females to choose less prestigious career paths and have little affect on males' career aspirations. The lower GPA's affect both sexes in reduced career aspirations, but still affects female students to a greater degree.
table of contents