null hypothesis
Hypothesis that there really is no relationship between our hypothesis variables in the population. Even if there is some relationship between the variables in our sample, the null hypothesis postulates that our findings can be attributed to sampling erro
direction hypothesis
A hypothesis that predicts which categories of the independent variable will have higher or lower values on the dependent variable.
directional observation
A way to operational define variables by observing actual behavior.
one- tailed test of significant
A significance test that places the entire critical region at the predicted end of the theoretical sampling distribution. There is no critical region at the the other end
level of significant
The cutoff point that separates the critical region probability from the rest of the area of the theoretical sampling distribution.
alpha level (rejection level)
The cutoff level (usually 0.05) separating statistically significant findings from non-significant findings. Also called rejection level.
statistical significant
What we have the probability that the null hypothesis is true, and is low enough to reject the null hypothesis as a plausible explanation for the relationship observed in a sample
percentile
A value that incorporates a certain percentage of rank ordered values in a distribution. For example, the 25th percentile is the value below which the lowest 25% of the value fall.
central limit theorm
An assumption that as the size of the sample increases, the theoretical sampling distribution based on that sample size will become increasingly normal in shape, and its mean will become increasingly close to the population mean.
frequency probability
The number of times that a particular event happens during repeated trials of some experiment.
sampling error
The possibility that the random variation(chance) can effect co-variation among variables in sample statistics.
probability value (-value)
The probability that a study's findings are attributable to sampling error.
confidence interval
A range of values constructed from sample data so that the population parameter is likely to occur within that range at a specified probability.
univariate analysis of variance
Statistical technique used to determine, on the basis of one dependent measure, whether samples are from populations with equal means.
multi-variance analysis of variance
An analysis of variance with more than one dependent variable.
meta-analysis
A study of studies, in which each study becomes a subject for the meta-analytical research study and advanced statistical procedures are used to merge the disparate findings of all the different studies
outlier
Very extreme values in a distribution of a variable.
quasi-experimental designs
Research method similar to an experimental design except that it makes use of naturally occurring groups rather than randomly assigning subjects to groups.
qualitative methods
Methods that attempt to collect information about the social world that cannot be readily converted to numeric form.
quantitative methods
Research methods that typically are used when studies aim to develop precise, objective, and generalizable findings. these studies rely on quantitative analysis: that is, they involve numbers and statistics.
ratio level of measurement
Measuring variables in such a way that there is a true zero point and difference between different levels have the same mathematical meanings. Number of arrests is an example of a variable that can be measured at the ratio level.
regression analysis
Developing an equation that estimates the value of a dependent variable based on the value of an independent variable
regression equation
The equation used in regression analysis to predict the value of one variable based on the value of another variable. Also called least squares regression equation.
relative frequencies
proportion or percentage of observations that fall into that category (used for categorical data). proportion equals the number of observations in a category divided by the total number of observations. percentage is the proportion multiplied by 100
research hypothesis
A testable, concise statement involving the expected relationship between two or more variables. Can be non-directional or directional
sample
a part of the population from which we have data
scatterplot
a graphed cluster of dots, each of which represents the values of two variables. The slope of the points suggests the direction of the relationship between the two variables. The amount of scatter suggests the strength of the correlation (little scatter i
selection bias
Also called Undercoverage, a bias that is introduced when the way a sample is selected systematically excludes some part of the population of interest. Results in a sample that is under-representative of a population.
skew distribution
A distribution in which more values fall on one side of the mean than on the other side of the mean. this imbalance creates a difference between the mean and the median.
Spearman's rho
A commonly used nonparametric formula for calculating correlation coefficients with variables that are at the ordinal level of measurement or with interval or ratio level data that are not distributed normally. It produces a correlation coefficient rangin
standard deviation
The most commonly cited measure of dispersion that refers to how far the scores in a distribution are deviating from the mean on an average. It is the square root of the variance.
standard error of the mean/standard error
Special type of standard deviation that measures validity in the sampling distribution. Provides us with a standard to describe the amount by which sample means deviate from the mean of their sampling distribution and their population mean. Errors = error
standardized beta
A statistic use in multiple regression analysis to depict the relative degree of influence a variable has in explaining the variation in the dependent variable when other variables are controlled. the larger the beta the greater the influence.
statistic
A branch of mathematics used by researchers to organize, summarize, and interpret data.
statistical conclusion validity
A desirable attribute of a study attained when the researchers use an appropriate statistical procedure and are correct in deciding whether to rule out chance as an alternative explanation for study findings.
statistically significant
Descriptive of a finding that allows us to rule out sampling error as a plausible explanation for the finding.
substantive significance
The practical value or importance of a relationship - that is, how meaningful it is to clients, significant others, society, or practitioners concerned about a problem
symmetrical
A characteristic of normal distributions in which the right and left halves of the curve are mirror images of each other.
theoretical sampling distribution
The distribution of outcomes produced by an infinite number of randomly drawn samples or random subdivisions of a sample. This distribution identifies the proportion of times that each outcome of a study could be expected to occur as a result of chance.
trimmed means
A mean that is calculated after trimming off outliers at both the high and low ends of the distribution when those outliers make up a very small percentage of the distribution (usually 5%
t-test
A significance test that can be used with an interval or ratio-level dependent variable and a dichotomous nominal level independent variable that has only two categories. the most common use of the two test in evaluating programs and practice is to compar
two factors (ANOVA)
An analysis of variance involving two nominal-level independent variables.
two-standard-deviation procedure
A simple approach that practitioners can use to approximate statistical significance ( as an alternative to the t-test) for outcome data at the interval or ratio level of measurement. this procedure involves calculating an effect size, using the standard
two-tailed test of significance
A significance test that divides the critical region at both ends of the theoretical sampling distribution.
type 1 error
An error that occurs whenever we reject a true null hypothesis.
type 2 error
An error that occurs when we fail to reject a false null hypothesis.
validity
whether a measure truly and accurately measure what it intends to measure.
variability
The amount of dispersion in the distribution of a particular variable
variables
Any measurable conditions, events, characteristics, or behaviors that are controlled or observed in a study.
variance
A measure of dispersion that is the average of the squared deviations from the mean.
variation
the extent of which the values in a distribution are clustered near one another or are scattered away from one another,
Wilcoxon T test
A nonparametric significance test that uses rank-ordered values to compare the difference between two related groups with ordinal data.
x-axis (abscissa)
A horizontal line in a graph that typically displays the values of a variable.
random sampling
Selecting a sample using random numbers to ensure that your biases, limited knowledge about a population, or errors in judgment cannot influence which elements are selected for inclusion in your study and which are not.
probability
The likelihood of a particular outcome occurring which is equal to the number of ways that particular outcome can occur divided by the total number of possible outcome.
probability sampling
Selecting a sample randomly so that every element in the larger population has an equal chance of being selected
frequency polygon (line graph)
A graph that uses single points instead of bars to convey the y-axis amount for each value along the x-axis.
frequency distribution
A list of the number or percentage of cases for each category of a variable.
hierarchical method
One method for multiple regression used when we hypothesize in advance which variables, or which sets of variables are more influential then others in predicting the dependent variable.
histogram
A bar graph for metric data with bars that touch each other.
independent variables
variables that are postulated to explain or cause something
inferential statistics
Inferential statistics allows us to make inferences about a larger group of individuals (a population) on the basis of data collected about a much smaller group (a sample).
internal validity
Experiments that have a high degree of convincing changes in behavior are a function of the independent variable and are not the result of uncontrolled or unknown variables
hypothesis
A tentative statement about the relationship between two or more variables
dependent variables
A factor that the experimenters measure to see if it is affected by the independent variable.
correlation coefficient
A statistic that depicts the strength of a correlation between two variables that are at the ordinal, interval or ratio level of measurement.
clinical statistic
Evaluating the effectiveness of clinical intervention.
coding
converting data to a machine readable format by assigning a code number or a code letter to each category of a variable.
chance (sampling error)
Refers to the possibility that random variation (sampling error) can effect co-variation among variables in sample statistics.
chi-square test of statistical significance
The most commonly used nonparametric test, which assesses the probability that sampling error explains the relationships we observed between nominal level variables displayed in cross-tabulation tables.
absolute frequency
a simple count of the number of cases per category of a variable.
cumulative frequencies
The sums of relative or absolute frequencies
data
Data consists of info coming from observations, counts, measurements, or responses.
descriptive statistic
Statistics that organize, summarize and display the data collected in a particular study without trying to develop inference beyond the sample or trying to rule out sampling error in hypothesis testing.
directional hypothesis
A hypotheses that makes a specific prediction about the direction of the relationship between two variables
confidence level
The estimate probability that a population parameter lies within a given confidence interval
correlation
the degree to which the values of two variables vary together in a consistent fashion.
Control group
those clients who do not receive the intervention being evaluated in an experiment.
interquartile range
the range for the middle 50% of values in a rank -ordered distribution
interval level of measurement
Measuring variables in such a way that differences between different levels have the same meanings. for example, the difference between an IQ of 95 and 100 is considered to be of the same magnitude as the difference between 100 and 105
level of measurement
Nominal, Ordinal, Interval, Ratio - Four levels of measurement each containing differing amount of information as follows: Nominal = category, Ordinal = rank, Interval = equal distance, Ratio = all plus true 0 point
measure of central tendency
Statistics that use a single number t summarize data the ordinal, interval, or ratio level of measurement. Measure of central tendency include the mean median and mode.
normal curve
A symmetrical, bell-shaped curve that describes the distribution of many types of data; most scores fall near the mean and fewer and fewer near the extremes.
one sample t-test
a t-test that is used when we want to compare a sample statistic to a population statistic.
metric measurement
Measurement that are at the ordinal, interval or ratio-level
one tailed test of significance
A significance test that place the entire critical region at the predicted end of the theoretical sampling distribution. There is no critical region at the other end
operation definitions
Research definitions that specify precisely what observations will determine what attribute or value category applies to a particular variable for a particular research participant. Thus we might operationally define level of substance abuse in terms of t
parameter
a summary statistic describing a given variable for an entire population
population
the entire universe of cases to which we seek to generalize from our sample data
nominal-level variables
Variables that vary only in categories that are qualitative in nature (such as gender).
non directional hypothesis
A predicted relationship that does not specify which attributes on one variable will have higher or lower value on the other variable.
bimodal
A distribution that contain two modes
bell shaped curve (bell curve)
Another name for a normal curve based on the fact that the normal distribution is curved to the shape of a bell
coefficient of determination
R squared, used to assess how well the regression model fits the data: represents the percentage of the variation in the dependent variable that is explained by the regressions model.
bar graph
Diagram depicting frequency distribution, using bars to show the number of percentage of cases for each categories of a nominal level variable.
critical region
The area of the theoretical sampling distribution where our sample statistic needs to fall in order to be deemed statistically significant (that is , too unlikely to be attributable to sampling error).
evidence based practice
Using the best scientific evidence available to guide you practice
group frequency distribution
A frequency distribution where scores are grouped into intervals rather than listed as individual values
attributes/value categories
The concepts that make up a variable. The attributes, or value categories, of gender, for example, are male or female.
range
The simplest measure of dispersion. It is the total number of possible values between the minimum and maximum values in a distribution.