Alphabetical listing links:
A
B
C
- COEFFICIENT OF DETERMINATION - the proportion of variability accounted for by a correlation and the square of the PEARSON CORRELATION COEFFICIENT
- CONFOUNDING VARIABLE - variable that covaries with an independent variable.
- CONSTANT of a LINEAR MODEL(see also INTERCEPT)
- CONTINGENCY TABLE - a table used to demonstrate the relationship between two or more variables, typically expressed as categories (see NOMINAL DATA).
- CONTINUOUS VARIABLE - variable with infinite number of possible values.
- CRITICAL SIGNIFICANCE LEVEL - the significance level selected for the testing of a hypothesis, usually denoted by the lowercase Greek letter alpha (α). If the calculated PROBABILITY calculated for the test is lower than the critical significance level, the result is unlikely to have occurred by chance and the NULL HYPOTHESIS is rejected.
- CRITICAL VALUE - value of a statistic that corresponds to a give level of probability.
D
- DEGREES OF FREEDOM - an adjusted sample size used in calculating various statistics eg., n-1.
- DEPENDENT VARIABLE - variable that depends on the independent variable.(see also INDEPENDENT VARIABLE).
- DESCRIPTIVE STATISTICS - statistics that describe data.
- DISCRETE VARIABLE - variable consisting of separate indivisible categories.
E
- ERROR BAR – T shaped line indicating variability of data on graphs.
- EXPECTED FREQUENCY DISTRIBUTION - frequency distribution seen in measured data. (see also FREQUENCY DISTRIBUTION)
F
- FREQUENCY - the number of times an observation of a particular value occurs.
- FREQUENCY DISTRIBUTION - number of individuals/items in each category or interval (graph or table).
- F-STATISTIC - the ratio of 'model' variance to residual variance. For example, in analysis of variance, F is calculated as the variance between treatments, divided by the variance within treatments.
G
H
I
- INDEPENDENT VARIABLE - variable that causes changes in dependent variable.(see also DEPENDENT VARIABLE)
- INFERENTIAL STATISTICS - statistics that allow us to draw inferences about populations from samples.
- INTERCEPT of a LINEAR MODEL - (see also CONSTANT)
- INTERQUARTILE RANGE – difference between 1st & 3rd quartile (a measure of variability and an example of a descriptive statistic).(see also DESCRIPTIVE STATISTICS)
J
K
L
- LINEAR MODEL - A model that describes the relationship between a DEPENDENT VARIABLE y and an INDEPENDENT VARIABLE x in the form y = bx + c where b is the SLOPE and c is the CONSTANT
- LINEAR RELATIONSHIP - relationship between two variables that can be described effectively by a LINEAR MODEL.
M
- MEAN - average (a measure of central tendency and an example of a descriptive statistic).(see also DESCRIPTIVE STATISTICS)
- MEDIAN - middle value (a measure of central tendency and an example of a descriptive statistic).(see also DESCRIPTIVE STATISTICS)
- MODE - most frequent value (a measure of central tendency and an example of a descriptive statistic)(see also DESCRIPTIVE STATISTICS)
N
- NOMINAL DATA - categorical data.
- NORMAL DISTRIBUTION - a very common and important symmetrical frequency distribution of continuous data.
- NULL HYPOTHESIS - a hypothesis used in statistical tests of predictions, where the hypothesis is usually that the observations have arisen by chance.
O
- OBSERVED FREQUENCY DISTRIBUTION - frequency distribution seen in measured data. (see also FREQUENCY DISTRIBUTION)
- ORDINAL DATA - ranked data.
P
- PARAMETER - a value which describes a population.
- PARAMETRIC CRITERIA - data should be (1) normally distributed, and (2) variances should be similar (the standard deviation of the most-variable sample should be less than 10 times the standard deviation of the least-variable sample). (see also NORMAL DISTRIBUTION, STANDARD DEVIATION and VARIANCE)
- PEARSON CORRELATION - parametric test for the significance of the LINEAR RELATIONSHIP between two variables.
- PERCENTAGE - a proportion expressed as a value relative to one hundred e.g. three-quarters = 0.75 = 75%.(see also PROPORTION)
- POPULATION - entire set of individuals/items of relevance to a research question.
- POPULATION SIZE (N) - number of individuals/items in a population.
- PREDICTION - statement of what pattern in data expected if research hypothesis is true.
- PROBABILITY - measure of the likelihood of an event or result. For instance, for a dice there is a probability of one in six that a particular face will turn up on a given throw - usually written as P = 0.16666. Results of statistical tests are usually expressed relative to a CRITICAL SIGNIFICANCE LEVEL, for instance P < 0.05 indicates that the null hypothesis has been rejected on the basis that the pattern observations had a probability of less than one in twenty of occurring by chance.
- PROPORTION - one value expressed relative to a second value, for example a fraction.(see also PERCENTAGE)
Q
R
- r - symbol given to the PEARSON CORRELATION COEFFICIENT.
- r2 - symbol given to the COEFFICIENT OF DETERMINATION, the proportion of variability accounted for by a correlation and the square of the PEARSON CORRELATION COEFFICIENT.
- RANDOM SAMPLING - selecting individuals from a population without bias.
- RANGE – difference between highest and lowest value (a measure of variability & an example of a descriptive statistic).
- RELATED DATA - if data are related, an observation of one variable can be linked to a corresponding observation of a second variable. Obviously, this means that there must be the same number of observations in each set of data(see also UNRELATED DATA)
- RESEARCH HYPOTHESIS - educated guess to a research question.
S
- SAMPLE - a subset of data from a population.
- SAMPLE SIZE (n) - number of individuals/items in a sample.
- SAMPLING DISTRIBUTION - frequency distribution of a statistic eg., mean.
- SAMPLING ERROR - discrepancy between a statistic and a population.
- SCALE DATA - ranked data with scale.
- SLOPE of a LINEAR MODEL - the increase in a dependent variable for a given increase in the independent variable.
- SPEARMAN CORRELATION - non-parametric test for the significance of the LINEAR RELATIONSHIP between two variables..
- STANDARD DEVIATION (s or σ) – square root of variance (a measure of variability & an example of a descriptive statistic).
- STANDARD ERROR OF THE MEAN - the standard deviation of the sampling distribution of means.
- STANDARDIZED NORMAL DISTRIBUTION – a normal distribution whose mean has been shift to zero & x axis marked in units of standard deviation (z scores).
- STATISTIC - a value which describes a sample.
- STRAIGHT LINE - see LINEAR MODEL or LINEAR RELATIONSHIP
- SUM OF SQUARES - sum of squared deviations.
T
U
- UNRELATED DATA - if data are related, an observation of one variable is not associated with a corresponding observation of a second variable (see also RELATED DATA.)
V
- VARIANCE (s2 or σ2) - mean squared deviation (a measure of variability and an example of a DESCRIPTIVE STATISTIC). Note for a SAMPLE this is the SUM OF SQUARES divided by the DEGREES OF FREEDOM.
W
X
Y
Z
- Z SCORES - a standardized measure of the distance of a score from the mean (see also STANDARDIZED NORMAL DISTRIBUTION).
others: numerical and non-English characters
- 95% CONFIDENCE LIMITS - the range around a sample mean with very good chance of including population mean.
- α (Greek lowercase alpha) - symbol usually given to CRITICAL SIGNIFICANCE LEVEL.
- Σ (Greek capital sigma) - sum of.