The SmithJohnson Manual
Bivariate Statistics: Considers two variables together and describes the relationship between variables (e.g. crosstabulation)
Closed-ended questions: Both asks a question and gives respondents specific answer categories to choose from.
Conjoint Analysis: Mail or intercept survey measuring the relative importance of a set of identified attributes that people may consider when choosing between competing candidates, products or services.Focus Group: Group discussion of 8-10 people led by a skilled moderator who directs the discussion to generate unforeseen hypotheses and probe for complicated perceptions and feelings. Also see Qualitative Analysis.
Intercept Survey: A questionnaire administered face-to-face to a sample of respondents as they pass by the interviewer.
Interval Data: Indicates a difference among categories, that the categories can be ranked, and specifies an equal distance between categories (e.g. IQ score-95, 110, 125…).
Multivariate Statistics: Analysis of the simultaneous relationships among several variables. Examples include logistic regression and factor analysis.
Nominal Data: Indicates that there is a difference among categories (e.g. gender).
Nonsampling Error: Errors other than those resulting from the measuring a sample rather than a population (e.g. measurement errors, data entry errors, question bias, sample bias).
Null Hypotheses: The statistical assumption that observed differences are due to chance (sampling error) and that the observed relationship between variables is due to chance.
Open-ended Question: An unstructured question allowing respondents to answer in their own words.
Ordinal Data: Indicates a difference among categories plus that the categories can be ordered and ranked (e.g. letter grades – A, B, C, D…).
Population: The complete set of individuals about whom the researcher is interested.
Probability Sample: Virtually every unit in the population has a known and non-zero probability of being included in the sample.
Qualitative Research: Consists of observations not easily reduced to numbers, such as a focus group.
Quantitative Research: Consist of information in numerical form, such as a telephone survey.
Sample: A subset of a population used to collect data.
Sample Survey: Research whereby a subset of the population is measured.
Sampling Error: The difference between a statistic and a parameter that is due to the use of the sample rather than the census.
Sampling Frame: The list (or quasi-list) of units from which a sample is selected (e.g. a telephone directory).
Statistical Significance: The conclusion that the apparent sample relationship between two variables is not due to chance (sampling error).
Strength of Association: The question of how well two variable are related (correlated). Statistical significance should not be confused with Strength of Association.
Univariate Statistics: Describes a single variable. An example is a frequency table or measure of central tendency.
Variance: A measure of the average squared variability, calculated by dividing variation by the number of scores ( or the number of scores minus one when a sample is used to estimate population).
Variation: A measure of total squared variability, calculated by summing across the squared difference between each score and mean.