Term
|
Definition
| A process of scientific reasoning in which systematic observation leads to the development of theory and hypotheses. |
|
|
Term
|
Definition
| Research to describe how the world is. Contrasts with causal research. |
|
|
Term
|
Definition
| Data collected by others, such as existing government surveys, administrative records, or transcripts. Contrasts with primary data. |
|
|
Term
|
Definition
| Collecting new data to provide a description or explanation of the world. Contrasts with secondary data. |
|
|
Term
|
Definition
| A process of scientific reasoning in which theories lead to hypotheses (predictions) that are compared with data (observation). |
|
|
Term
| Contingent data, limited to certain times, places, or contexts. |
|
Definition
| Unprocessed or unaggregated observations—raw data. |
|
|
Term
|
Definition
| The extent to which the results of a study project to a wider group or context of interest. Also called external validity. |
|
|
Term
|
Definition
| The search for published sources describing the results of research or information provided by others. |
|
|
Term
|
Definition
| The original collection or analysis of data to answer a new research question or to produce new knowledge. |
|
|
Term
|
Definition
| Theory of knowledge, or ways of knowing. |
|
|
Term
|
Definition
| Research to answer “What if?” questions. Contrasts with descriptive research. |
|
|
Term
|
Definition
| The process in which an organization collects information to measure and report on how well it is doing, usually with the goal of managing to improve its performance. |
|
|
Term
|
Definition
| Process in which studies or proposals are reviewed and approved (or rejected) by a group of peers—other researchers in the same field—who render a judgment on the methodology and worth of the paper or proposal. |
|
|
Term
|
Definition
| A detailed causal process that produces an outcome. Called mechanism for short. |
|
|
Term
|
Definition
| A relationship (correlation or association) due to causation. When one variable changes, another one changes as a consequence. Contrasts with noncausal relationship. |
|
|
Term
| Cross-sectional variation |
|
Definition
| Variation across individuals, organizations, or places at a single point in time. |
|
|
Term
|
Definition
| Models that describe relationships between variables at different units of analysis. Also called multilevel models. |
|
|
Term
|
Definition
| The variable whose effect is of interest—the cause, the explanatory variable, the treatment, or the predictor. Contrasts with dependent variable. |
|
|
Term
|
Definition
| The financial, human, and material resources required by the program described in more implementation-oriented logic models. See also activities, outputs, and outcomes. |
|
|
Term
|
Definition
| A variable along a causal pathway. Also called mediator. |
|
|
Term
|
Definition
|
|
Term
|
Definition
| A diagram or equation or other representation that serves to articulate and communicate a theory. |
|
|
Term
|
Definition
| A variable that changes the magnitude of the effect of another variable. See also interaction. |
|
|
Term
|
Definition
| A relationship (correlation or association) between two variables in which neither is a direct cause of the other. Contrasts with causal relationship. See also spurious relationship. |
|
|
Term
|
Definition
| Immediate products of activities, such as people trained, brochures distributed, or citations issued, described in implementation-oriented logic models. See also activities, inputs, and outcomes. |
|
|
Term
|
Definition
| The response of interest to a program. See also activities, inputs, outputs, and dependent variable. |
|
|
Term
| Positive (+) relationship |
|
Definition
| Relationship in which the two related variables move in the same direction. Contrasts with negative relationship. |
|
|
Term
| Positive (statement or theory) |
|
Definition
| A statement or theory that describes how things really are. Contrasts with normative. |
|
|
Term
| Normative (statement or theory) |
|
Definition
| A statement or theory that describes how things should be. Contrasts with positive. |
|
|
Term
|
Definition
| A logical description of how a particular corner or aspect of the world works. |
|
|
Term
|
Definition
| The objects, individuals, or things described by the variables or theory. |
|
|
Term
|
Definition
| Research that focuses on a single complex case, applying multiple qualitative and sometimes quantitative methods. |
|
|
Term
|
Definition
| An individual, group, or institution that is the focus of qualitative research. |
|
|
Term
|
Definition
| The analysis of the content of qualitative data, such as texts or images, usually involving a process of coding. |
|
|
Term
|
Definition
| Qualitative research method to describe and understand the culture of a group of people, often using participant observation. |
|
|
Term
|
Definition
| A qualitative group interviewing procedure that involves typically 6 to 12 participants, seated around a table, and a moderator who asks questions and guides the discussion. |
|
|
Term
|
Definition
| A theory that emerges from observations made in a qualitative study and is grounded in the specific setting. |
|
|
Term
|
Definition
| The notion that language allows us to stand in someone else’s shoes and see the world from the other’s perspective. |
|
|
Term
|
Definition
| A set of open-ended questions, sometimes accompanied by probes, that help guide or structure the discussion in a semistructured interview. |
|
|
Term
|
Definition
| Survey using more than one mode of data collection. |
|
|
Term
| Moderator (of a focus group) |
|
Definition
| The person who guides a focus group. Also called a facilitator. |
|
|
Term
|
Definition
| A question that cannot be answered with a limited set of possible answers and gives the person answering the opportunity to choose what information to provide. Contrasts with closed-ended question. |
|
|
Term
|
Definition
| A qualitative research method in which the researcher participates in and observes his or her subjects. |
|
|
Term
|
Definition
| Subjects or cases of research are chosen for a purpose, not to provide a sample that is representative of a population, for qualitative research. See also theoretical sampling. |
|
|
Term
|
Definition
| The raw data from qualitative research, which can take the form of field notes, interview transcriptions, video or audio recordings, or documents, among others. |
|
|
Term
|
Definition
| Research that involves language, images, and other forms of expressing meaning that researchers then interpret—research that does not involve numbers or quantification. |
|
|
Term
|
Definition
| The point in the process of collecting qualitative data when few new issues or questions arise that have not already been discovered. |
|
|
Term
|
Definition
| A qualitative research method that involves interviewing with an interview guide, including a planned set of open-ended questions. |
|
|
Term
|
Definition
| A qualitative research method that involves interviewing with no predetermined set of questions. |
|
|
Term
|
Definition
| Systematic error (distortion) in the measurement of some trait or construct. Also called systematic measurement error. |
|
|
Term
|
Definition
| Variables that refer to categories and not to actual quantities. Contrasts with quantitative variables. |
|
|
Term
|
Definition
| Stage in the measurement process in which the construct (or concept) to be measured is carefully and fully defined. |
|
|
Term
|
Definition
| The extent to which a measure concurs or agrees with other established classifications or measures taken at the same time. |
|
|
Term
|
Definition
| The concept or trait that a measure is trying to capture. Also called trait. |
|
|
Term
|
Definition
| The extent to which a measure correlates with other variables or behaves in a statistical model in a way that would be expected, based on theory and prior research. |
|
|
Term
|
Definition
| How well a measure captures all the important dimensions of a construct. |
|
|
Term
|
Definition
| The extent to which a measure correlates with other closely related measures in the same data set. |
|
|
Term
| Criterion-related validity |
|
Definition
| The extent to which a measure relates, empirically, to various criteria that can demonstrate its validity. |
|
|
Term
|
Definition
| Extent to which a measure is independent of (not correlated with) other measures in the same data set that it does not logically relate to. |
|
|
Term
|
Definition
| On the face of it, how well a measure captures what it is supposed to measure—the extent to which a measure makes intuitive sense. |
|
|
Term
|
Definition
| A composite measure composed of multiple items, which may be selected for different reasons. |
|
|
Term
|
Definition
| Some observable measure that reveals information about a factor or latent trait. |
|
|
Term
|
Definition
| Refers to the intercorrelation of the items or indicators of a multi-item scale. See also Cronbach’s alpha and split-half reliability. |
|
|
Term
|
Definition
| How similar the scores of different raters or interviewers are when they measure the same person or object. |
|
|
Term
|
Definition
| A construct or trait that is not directly observable. Contrasts with manifest construct. |
|
|
Term
|
Definition
| The distinction between quantitative and categorical variables, or “ladder of measurement”: nominal, ordinal, interval, and ratio. |
|
|
Term
|
Definition
| A construct or trait that is directly observable. Contrasts with latent construct. |
|
|
Term
|
Definition
| The score or result produced by a measurement process. |
|
|
Term
| Nominal categorical variables |
|
Definition
| Categorical variables that have no intrinsic order. Contrasts with ordinal categorical variables. |
|
|
Term
| Ordinal categorical variables |
|
Definition
| Categorical variables that can be put in a meaningful order. Contrasts with nominal categorical variables. |
|
|
Term
|
Definition
| The process in which an organization collects information to measure and report on how well it is doing, usually with the goal of managing to improve its performance. |
|
|
Term
|
Definition
| The extent to which a measure predicts logically related outcomes or behaviors in the future. |
|
|
Term
|
Definition
| Carefully specified procedures for using the instruments properly in measurement. |
|
|
Term
|
Definition
| A measure that substitutes for another unavailable measure. |
|
|
Term
|
Definition
| A person who responds to a survey providing information about someone else. |
|
|
Term
|
Definition
| The field that deals with the measurement of latent traits or constructs, primarily using instruments such as scales or tests composed of multiple items or questions, and related statistical techniques. |
|
|
Term
|
Definition
| Variables that take the form of numbers that refer to actual quantities of something. Contrasts with categorical variables. |
|
|
Term
|
Definition
| Errors—deviations from the true construct in a measure—that are not systematic and average out to zero. Also called noise. Contrasts with bias, systematic measurement error. |
|
|
Term
|
Definition
| Consistency of a measure. |
|
|
Term
|
Definition
| A composite measure composed of multiple items and thought to reflect a single latent construct. |
|
|
Term
|
Definition
| An approach to determining reliability based on measuring the same thing twice. |
|
|
Term
|
Definition
| Concept, construct, or characteristic of which the measurement is sought. See also construct. |
|
|
Term
|
Definition
| How well a measure represents the construct of interest. |
|
|
Term
|
Definition
| Every member of a population. |
|
|
Term
|
Definition
| A probability sampling method in which more aggregated units (clusters) are sampled before sampling individuals. |
|
|
Term
|
Definition
| Probability sampling methods that are more complex than simple random sampling, such as cluster sampling, stratified sampling, and disproportionate sampling. |
|
|
Term
|
Definition
| A range of values in which we have a defined level of confidence (e.g., 95%) that the true value of the statistic being estimated lies. See also margin of error. |
|
|
Term
|
Definition
| Share who are reached from those sampled from the sampling frame. See also cooperation rate and response rate. |
|
|
Term
|
Definition
| A nonprobability sample that was chosen for reasons of convenience, such as proximity. |
|
|
Term
|
Definition
| Share who cooperate with a survey request from among those contacted. See also contact rate and response rate. |
|
|
Term
|
Definition
| Bias in survey that occurs when members of the sampling frame are systematically different from the target population in a way related to the measures. |
|
|
Term
|
Definition
| The loss of (or gain in) precision due to a particular complex survey sampling design. See also effective sample size. |
|
|
Term
| Disproportionate sampling |
|
Definition
| A variation on stratified sampling in which some strata are sampled at different rates. Also called oversampling. |
|
|
Term
|
Definition
| The comparable sample size from a simple random sample; it expresses the design effect (often a loss) due to complex sampling. See also design effect. |
|
|
Term
|
Definition
| The extent to which the results of a study project to a wider group or context of interest. Also called external validity. |
|
|
Term
|
Definition
| Using samples to learn about the population, or using evidence to identify a causal relationship. |
|
|
Term
|
Definition
| A method for pooling together multiple smaller studies to get a much bigger, combined study. |
|
|
Term
|
Definition
| A probability sampling method in which more aggregated units (clusters) are sampled and then sampling occurs within the aggregates. |
|
|
Term
|
Definition
| Bias in survey results that occurs when those who do not respond are systematically different from those who do respond in a way related to what the survey aims to measure. |
|
|
Term
|
Definition
| The population the study aims to investigate. Also called universe. |
|
|
Term
|
Definition
| Refers to the amount of random variability in the results of a sample (the less variability, the more precision). See also sampling variability. |
|
|
Term
| Probability proportional to size (PPS) sampling (Proportionate Sampling) |
|
Definition
| Sampling in which elements, such as businesses or communities, are selected with a probability proportional to their size. |
|
|
Term
|
Definition
| A method of sampling that uses chance to select people (or elements) from a population. Also called random sampling. |
|
|
Term
|
Definition
| Likelihood of responding to a survey or survey question. |
|
|
Term
|
Definition
| Subjects or cases of research are chosen for a purpose, not to provide a sample that is representative of a population, for qualitative research. See also theoretical sampling. |
|
|
Term
| Random digit dialing (RDD) |
|
Definition
| A telephone survey method that gives both listed and unlisted numbers an equal chance of being selected by replacing random digits at the ends of listed residential telephone numbers. |
|
|
Term
|
Definition
| Share who respond to a survey from among those sampled from a sampling frame. See also contact rate and cooperation rate. |
|
|
Term
|
Definition
| A subset of people or elements selected from a population. Contrasts with census. |
|
|
Term
|
Definition
| Process of selecting people or elements from a population for inclusion in a research study. |
|
|
Term
|
Definition
| When results of a study differ systematically from the population because of shortcomings in the sampling process. See also coverage bias and nonresponse bias. |
|
|
Term
|
Definition
| Error in sample statistics due to random chance of who ends up in a sample. See also sampling variability. |
|
|
Term
|
Definition
| A list, map, or other representation of a population used for purposes of drawing a sample. |
|
|
Term
|
Definition
| Variability in sample statistics, across different samples, due to random chance of who ends up in a sample. See also sampling error. |
|
|
Term
|
Definition
| Selecting of people (or elements) from a population in such a way that each individual has an equal chance, or probability, of selection. |
|
|
Term
|
Definition
| Method of sampling or finding study subjects in which interviewees are asked to refer people they know to the researcher for inclusion in the sample. See also respondent-driven sampling. |
|
|
Term
|
Definition
| The precision of the estimate—how good a job we expect it to do, on average. |
|
|
Term
|
Definition
| Probability sampling method in which a sample is drawn separately from each group—each stratum—and the population is divided into exhaustive and mutually exclusive strata. See also strata. |
|
|
Term
|
Definition
| Probability sampling method in which individuals or elements are sampled at even intervals—every kth individual for some integer k. |
|
|
Term
|
Definition
| Bias in a study that occurs when volunteers differ from a more representative sample of the population in ways that influence the findings of the study. |
|
|
Term
|
Definition
| Measurements of many individuals, organizations, or places at a single point in time. |
|
|
Term
|
Definition
| A place where data and related documentation from various surveys and studies are stored and made publicly available for research. |
|
|
Term
|
Definition
|
|
Term
|
Definition
| Data providing information about the data. |
|
|
Term
|
Definition
| Data at its most basic level of observation—or unit of analysis—often of individual people or households. Contrasts with aggregate data and ecological data. |
|
|
Term
|
Definition
| Not containing information that identifies who an individual is or enables identification through indirect means. |
|
|
Term
|
Definition
| Repeated measures on the same individuals (or group or entity) over time. |
|
|
Term
|
Definition
| Repeated independent cross sections over time. |
|
|
Term
|
Definition
| Group of study subjects followed longitudinally over time. See also panel study. |
|
|
Term
|
Definition
| The raw data from qualitative research, which can take the form of field notes, interview transcriptions, video or audio recordings, or documents, among others. |
|
|
Term
|
Definition
| Completely aggregated or single-measure data over time. |
|
|
Term
|
Definition
| The objects (what or who) being described by the data being collected. See unit of analysis. |
|
|
Term
|
Definition
| When the researchers themselves do not know the identities of people in their study. Contrasts with confidentiality. |
|
|
Term
|
Definition
| Question that may be answered only with a limited set of predetermined response categories. Contrasts with open-ended question. |
|
|
Term
| Computer-assisted self-interviewing (CASI) |
|
Definition
| Interview provided to subjects on a computer so that software controls the flow of questions and data are entered in electronic form. |
|
|
Term
| Computer-assisted personal interviewing (CAPI) |
|
Definition
| In-person interviewing done on a computer so that software controls the flow of questions and data are entered in electronic form. |
|
|
Term
| Computer-assisted telephone interviewing (CATI) |
|
Definition
| Telephone interviewing done on a computer so that software controls the flow of questions and data are entered in electronic form. |
|
|
Term
|
Definition
| When the identity of people in a study is known to researchers but not revealed in publications or presentations of the study’s findings. Contrasts with anonymity. |
|
|
Term
|
Definition
| Mode of survey research in which self-administered questionnaires are distributed in group settings. |
|
|
Term
| Household interview survey |
|
Definition
| Survey conducted by visiting and interviewing people in their homes. |
|
|
Term
| Intercept interview survey |
|
Definition
| Mode of survey research in which people are stopped (intercepted) in public places and asked questions. |
|
|
Term
|
Definition
| When responses to an interview are influenced by the interviewer’s gender, ethnicity, age, and other characteristics. |
|
|
Term
| Mail self-administered survey |
|
Definition
| Mode of survey research in which forms are mailed to respondents along with instructions to complete and return the forms. |
|
|
Term
|
Definition
| Survey using more than one mode of data collection. |
|
|
Term
|
Definition
| A question that cannot be answered with a limited set of possible answers and gives the person answering the opportunity to choose what information to provide. Contrasts with closed-ended question. |
|
|
Term
|
Definition
| Survey in which the same respondents are tracked and repeatedly surveyed over time, sometimes over many years. See panel data. |
|
|
Term
|
Definition
| The tendency of respondents to a survey or interview to provide answers that make themselves look good or that are socially acceptable. |
|
|
Term
|
Definition
| The method or modality used to administer the questionnaire and collect the survey data, such as telephone, mail, or Internet. |
|
|
Term
| Telephone interview survey |
|
Definition
| Survey conducted by telephone, usually based on random digit dialing. |
|
|
Term
|
Definition
| A method of data collection that involves training research workers to systematically observe and record conditions or behaviors, typically using an observation rating form. |
|
|
Term
|
Definition
| Mode of survey research that uses web-based forms or questionnaires to gather responses. Also called Internet survey or online survey. |
|
|
