From PsychWiki - A Collaborative Psychology Wiki
Definition: a dependent variable is the variable that is measured
Application: In this study, scientists were testing the how diet effects fertility. The variable being measured here is fertility.
Definition: a nominal variables are mutually exclusive and exhaustive catagories differing in some qualitative aspect.
Application: In this article they are observing the pain tolerance of people with red hair. The variable "hair color" is a nominal variable as it is a qualitative measurement involving different catagories (red hair and non-red hair).
Definition: an ordinal variable or scale has the properties of a nominal scale, but in addition the observations may be ranked in order of magnitude (with nothing implied about the difference between adjacent steps on the scale)
Application: In this article they are ranking high schools in the US. They are listed from best rated to the lowest (listed in order) but the differences between each number are not equal.
Definition: has all the properties of an ordinal scale, and a given distance between measures has the same meaning anywhere on the scale
Application: This chart illustrates the ranking of GPAs from law schools across the country. The distance between each measure is the same throughout and since these are for law school students one can assume the absence of a 0.0 GPA.
Definition: has all the properties of an interval scale plus an absolute zero
Application: The above chart shows the average weight and length of the fetus during pregnancy. Here a measure of 0 on both scales would mean the absence of a fetus, or not pregnant.
Frequency Distribution (regular, grouped, relative, or cummulative)
Definition: shows the number of observations for the possible categories or score values in a set of data.
Application: Table 1 (scroll down past references) displays a frequency distribution (includes both regular and cummulative) showing the number of reported foodborne disease outbreaks, cases and deaths in the United States between 1998 and 2002.
Percentile (percentile or percentile rank)
Definition: the percentage of cases in a distribtion that falls below a given point on the measurement scale
Application: This article discusses how percentile rankings of law schools has impacted the admissions process. Due to the increased importance of the schools ranking they have reformatted their addmissions process to maintain the highest ranking.
Definition: a graph that consists of a series of connected dots about the midpoint of each possible class interval (height of the dots corresponds to frequency or relative frequency).
Application: This article and accompany frequency polygraph describes how the corn prices have changed relative to those of petroleum products over the past 40 years.
Definition: is used for qualitative data, it is a graph that is similar to a histogram, except that a space appears between the rectangles.
Application: Figure 2 is a visual representation of the percentage of female substance abuse treatment admissions, aged 12 to 17, categorized by age and primary substance of abuse.
Definition: used for qualitative date, the area in any piece of the pie shows the relative frequency of a category.
Application: Figure 1 in this report provides a visual representation (pie chart) of the percentage of female adolescent substance abuse treatment admissions, aged 12 to 17, categorized by their Race/Ethnicity.
Definition: the sum of all the scores divided by the total number of scores.
Applications: This article examines the rising amount of debt college students have when they graduate from college. In 2008 the average (mean) was $23,200.
Definition: the value that divides the distribution into halves.
Application: This article examines the change in median home prices within Southern California over a two year period.
Definition: the score that appears with the greatest frequency.
Application: In this study scientists tested the DNA from a sample group of 40 homosexual brothers and discovered the appearance of five different patches of genetic material grouped around a particular area on the X chromosome in 33 of them. In this instance the genetic material discovered would be the mode as it was found to be the highest number of similar DNA features amongst the sample group.
Definition: the mean of the squares of the deviation scores.
Examples: http://www.psychologytoday.com/blog/the-scientific-fundamentalist/200811/the-50-0-50-rule-in-action-sociosexual-orientation-and-ris http://www.psychologytoday.com/blog/the-scientific-fundamentalist/200809/the-50-0-50-rule-why-parenting-has-virtually-no-effect-chi
Application: In both of the above articles Satoshi Kanazawa, an evolutionary psychologist, explains his theory of the 50-0-50 rule of variance for personality and behavioral traits.
Definition: the square root of the deviance
Application: This study uses a standard deviation of 3 or more to define morbid obesity in children. That means all children 3 or more standard deviations from the mean are assumed to be morbidly obese.
Standard Scores (z-scores)
Definition: all have a fixed mean and fixed standard deviation, a "z-score" is one type and states how far away a score is from the mean in standard deviation units.
Application: This study was testing ACE inhibition for infants with single ventricle congenital heart defects. The only difference was an approximately 0.5 lower z-score for head circumference in the enalapril group, while the placebo group had a slightly above average z-score.
Definition: a graph of the bivariate distribution consisting of dots at the point of intersection of paired scores
Application: In this article a sports writer used a scatter plot to show when the first goals are scored in each game. He used this data to analyze his theory that slow starts have no effect on New Jersey Devils' success.
Definition: a measure of the degree of relationship between two variables
Application: In this study they are testing the correlation between watching movies with smoking scenes and the effects it has on Mexican-Americans experimenting with smoking. They showed that the more movies they watched with smoking scenes the more likely they were to smoke. This shows a positive correlation.
EXTRA CREDIT: Correlation does not equal causation
This article discusses the research findings of Gergory Paul, he is trying to prove that popular religious belief is caused by dysfunctional social conditions. While he was able to find a strong positive correlation between religious belief and dysfunction social conditions (particularly in the US) that does not prove one causes the other. More research needs to be done to elimate possible confounds. While we can not link the correlation to causation, it is interesting to consider his findings.
◄ Back to Descriptive Statistics PSY 202 Fall 2009 page