Research Methods for Social Workers

Abstracting indexing services Providers of specialized reference tools that make it possible to find information quickly and easily, usually through subject headings and/or author approaches.

Abstracts Reference materials consisting of citations and brief descriptive summaries from positivist and interpretive research studies.

Accountability A system of responsibility in which program administrators account for all program activities by answering to the demands of a programs stakeholders and by justifying the programs expenditures to the satisfaction of its stakeholders.

Aggregated case-level evaluation designs The collection of a number of case-level evaluations to determine the degree to which a program objective has been met.

Aggregate-level data Derived from micro-level data, aggregate-level data are grouped so that the characteristics of individual units of analysis are no longer identifiable; for example, the variable, gross national income, is an aggregation of data about individual incomes.

Alternate-forms method A method for establishing reliability of a measuring instrument by administering, in succession, equivalent forms of the same instrument to the same group of research participants.

Alternative hypothesis See Rival hypothesis.

Analytical memos Notes made by the researcher in reference to interpretive data that raise questions or make comments about meaning units and categories identified in a transcript.

Analytic generalization The type of generalizability associated with case studies; the research findings of case studies are not assumed to fit another case no matter how apparently similar; rather, research findings are tested to see if they do in fact fit; used as working hypotheses to test practice principles.

Annual report A detailed account or statement describing a programs processes and results over a given year; usually produced at the end of a fiscal year.

Antecedent variable A variable that precedes the introduction of one or more dependent variables.

Antiquarianism An interest in past events without reference to their importance or significance for the present; the reverse of presentism.

A Phase In case-level evaluation designs, a phase (A Phase) in which the baseline measurement of the target problem is established before the intervention (B Phase) is implemented.

Applied research approach A search for practical and applied research results that can be utilized in actual social work practice situations; complementary to the pure research approach.

Area probability sampling A form of cluster sampling that uses a three-stage process to provide the means to carry out a research study when no comprehensive list of the population can be compiled.

Assessment-related case study A type of case study that generates knowledge about specific clients and their situations; focuses on the perspectives of the studys participants.

Audit trail The documentation of critical steps in an interpretive research study that allows for an independent reviewer to examine and verify the steps in the research process and the conclusions of the research study.

Authority The reliance on authority figures to tell us what is true; one of the ways of knowing.

Availability sampling See Convenience sampling.

Axes Straight horizontal and vertical lines in a graph upon which values of a measurement, or the corresponding frequencies, are plotted.

Back-translation The process of translating an original document into a second language, then having an independent translator conduct a subsequent translation of the first translation back into the language of origin; the second translation is then compared with the original document for equivalency.

Baseline A period of time, usually three or four data collection periods, in which the level of the clients target problem is measured while no intervention is carried out; designated as the A Phase in single-system designs (case-level designs).

Between research methods approach Triangulation by using different research methods available in both the interpretive and the positivist research approaches in a single research study.

Bias Not neutral; an inclination to some form of prejudice or preconceived position.

Biased sample A sample unintentionally selected in such a way that some members of the population are more likely than others to be picked for sample membership.

Binomial effect size display (BESD) A technique for interpreting the r value in a meta-analysis by converting it into a 2 by 2 table displaying magnitude of effect.

Biography Tells the story of one individuals life, often suggesting what the persons influence was on social, political, or intellectual developments of the times.

B Phase In case-level evaluation designs, the intervention phase, which may, or may not, include simultaneous measurements.

Case The basic unit of social work practice, whether it be an individual, a couple, a family, an agency, a community, a county, a state, or a country.

Case-level evaluation designs Designs in which data are collected about a single client systeman individual, group, or communityin order to evaluate the outcome of an intervention for the client system; a form of appraisal that monitors change for individual clients; designs in which data are collected about a single client systeman individual, group, or communityin order to evaluate the outcome of an intervention for the client system; also called single-system research designs.

Case study Using research approaches to investigate a research question or hypothesis relating to a specific case; used to develop theory and test hypotheses; an in-depth form of research in which data are gathered and analyzed about an individual unit of analysis, person, city, event, society, etc.; it allows more intensive analysis of specific details; the disadvantage is that it is hard to use the results to generalize to other cases.

Categories Groupings of related meaning units that are given one name; used to organize, summarize, and interpret qualitative data; categories in an interpretive study can change throughout the data analysis process, and the number of categories in a given study depends upon the breadth and depth the researcher aims for in the analysis.

Category In an interpretive data analysis, an aggregate of meaning units that share a common feature.

Category saturation The point in a qualitative data analysis when all identified meaning units fit easily into the existing categorization scheme and no new categories emerge; the point at which first-level coding ends.

Causality A relationship of cause and effect; the effect will invariably occur when the cause is present.

Causal relationship A relationship between two variables for which we can state that the presence of, or absence of, one variable determines the presence of, or absence of, the other variable.

CD-ROM sources Computerized retrieval systems that allow searching for indexes and abstracts stored on compact computer discs (CDs).

Census data Data from the survey of an entire population in contrast to a survey of a sample.

Citation A brief identification of a reference that includes name of author(s), title, source, page numbers, and year of publication.

Classical experimental design An explanatory research design with randomly assigned experimental and control groups in which the dependent variable is measured before and after the treatment (the independent variable) for both groups, but only the experimental group receives the treatment (the dependent variable).

Client system An individual client, a couple, a family, a group, an organization, or a community that can be studied with case- and program-level evaluation designs and with positivist and interpretive research approaches.

Closed-ended questions Items in a measuring instrument that require respondents to select one of several response categories provided; also known as fixed-alternative questions.

Cluster diagram An illustration of a conceptual classification scheme in which the researcher draws and labels circles for each theme that emerges from the data; the circles are organized in a way to depict the relationships between themes.

Cluster sampling A multistage probability sampling procedure in which the population is divided into groups (or clusters) and the groups, rather than the individuals, are selected for inclusion in the sample.

Code The label assigned to a category or theme in a qualitative data analysis; shortened versions of the actual category or theme label; used as markers in a qualitative data analysis; usually no longer than eight characters in length and can use a combination of letters, symbols, and numbers.

Codebook A device used to organize qualitative data by applying labels and descriptions that draw distinctions between different parts of the data that have been collected.

Coding (1) In data analysis, translating data from respondents onto a form that can be read by a computer; (2) In interpretive research, marking the text with codes for content categories.

Coding frame A specific framework that delineates what data are to be coded and how they are to be coded in order to prepare them for analyses.

Coding sheets In a literature review, a sheet used to record for each research study the complete reference, research design, measuring instrument(s), population and sample, outcomes, and other significant features of the study.

Cohort study A longitudinal survey design that uses successive random samples to monitor how the characteristics of a specific group of people, who share certain characteristics or experiences (cohorts), change over time.

Collaterals Professionals or staff members who serve as indigenous observers in the data collection process.

Collective biographies Studies of the characteristics of groups of people who lived during a past period and had some major factor in common.

Collectivist culture Societies that stress interdependence and seek the welfare and survival of the group above that of the individual; collectivist cultures are characterized by a readiness to be influenced by others, preference for conformity, and cooperation in relationships.

Comparative rating scale A rating scale in which respondents are asked to compare an individual person, concept, or situation, to others.

Comparative research design The study of more than one event, group, or society to isolate explanatory factors; there are two basic strategies in comparative research: (1) the study of elements that differ in many ways but that have some major factor in common, and (2) the study of elements that are highly similar but different in some important aspect, such as modern industrialized nations that have different health insurance systems.

Comparison group A nonexperimental group to which research participants have not been randomly assigned for purposes of comparison with the experimental group. Not to be confused with control group.

Comparison group posttest-only design A descriptive research design with two groups, experimental and comparison, in which the dependent variable is measured once for both groups, and only the experimental group receives the treatment (the independent variable).

Comparison group pretest-posttest design A descriptive research design with two groups, experimental and comparison, in which the dependent variable is measured before and after the treatment for both groups, but only the experimental group receives the treatment.

Compensation Attempts by researchers to compensate for the lack of treatment for control group members by administering it to them; a threat to internal validity.

Compensatory rivalry Motivation of control group members to compete with experimental group members; a threat to internal validity.

Completeness One of the four criteria for evaluating research hypotheses.

Complete observer A term describing one of four possible research roles on a continuum of participant observation research; the complete observer acts simply as an observer and does not participate in the events at hand.

Complete participant The complete participant is at the far end of the continuum from the complete observer in participant observation research; this research role is characterized by total involvement.

Comprehensive qualitative review A nonstatistical synthesis of representative research studies relevant to a research problem, question, or hypothesis.

Computerized retrieval systems Systems in which abstracts, indexes, and subject bibliographies are incorporated in computerized data bases to facilitate information retrieval.

Concept An understanding, an idea, or a mental image; a way of viewing and categorizing objects, processes, relations, and events.

Conceptual classification system The strategy for conceiving how units of qualitative data relate to each other; the method used to depict patterns that emerge from the various coding levels in qualitative data.

Conceptual framework A frame of reference that serves to guide a research study and is developed from theories, findings from a variety of other research studies, and the authors personal experiences and values.

Conceptualization The process of selecting the specific concepts to include in positivist and interpretive research studies.

Conceptual validity See Construct validity.

Concurrent validity A form of criterion validity that is concerned with the ability of a measuring instrument to predict accurately an individuals status by comparing concurrent ratings (or scores) on one or more measuring instruments.

Confidentiality An ethical consideration in research whereby anonymity of research participants is safeguarded by ensuring that raw data are not seen by anyone other than the research team and that data presented have no identifying marks.

Confounding variable A variable operating in a specific situation in such a way that its effects cannot be separated; the effects of an extraneous variable thus confound the interpretation of a research studys findings.

Consistency Holding steadfast to the same principles and procedures in the qualitative data analysis process.

Constant A concept that does not vary and does not change; a characteristic that has the same value for all research participants or events in a research study.

Constant comparison A technique used to categorize qualitative data; it begins after the complete set of data has been examined and meaning units identified; each unit is classified as similar or different from the others; similar meaning units are lumped into the same category and classified by the same code.

Constant error Systematic error in measurement; error due to factors that consistently or systematically affect the variable being measured and that are concerned with the relatively stable qualities of respondents to a measuring instrument.

Construct See Concept.

Construct validity The degree to which a measuring instrument successfully measures a theoretical construct; the degree to which explanatory concepts account for variance in the scores of an instrument; also referred to as conceptual validity in meta-analyses.

Content analysis A data collection method in which communications are analyzed in a systematic, objective, and quantitative manner to produce new data.

Content validity The extent to which the content of a measuring instrument reflects the concept that is being measured and in fact measures that concept and not another.

Contextual detail The particulars of the environment in which the case (or unit of analysis) is embedded; provides a basis for understanding and interpreting case study data and results.

Contradictory evidence Identifying themes and categories that raise questions about the conclusions reached at the end of qualitative data analysis; outliers or extreme cases that are inconsistent or contradict the conclusions drawn from qualitative data; also called negative evidence.

Contributing partner A social work role in which the social worker joins forces with others who perform different roles in positivist and interpretive research studies.

Control group A group of randomly assigned research participants in a research study who do not receive the experimental treatment and are used for comparison purposes. Not to be confused with comparison group.

Control variable A variable, other than the independent variable(s) of primary interest, whose effects we can determine; an intervening variable that has been controlled for in the studys research design.

Convenience sampling A nonprobability sampling procedure that relies on the closest and most available research participants to constitute a sample.

Convergent validity The degree to which different measures of a construct yield similar results, or converge.

Correlated variables Variables whose values are associated; values of one variable tend to be associated in a systematic way with values in the others.

Cost-benefit analysis An analytical procedure that not only determines the costs of the program itself but also considers the monetary benefits of the programs effects.

Cost-effectiveness analysis An analytical procedure that assesses the costs of the program itself; the monetary benefits of the programs effects are not assessed.

Cover letter A letter to respondents or research participants that is written under the official letterhead of the sponsoring organization and describes the research study and its purpose.

Credibility The trustworthiness of both the steps taken in qualitative data analysis and the conclusions reached.

Criterion validity The degree to which the scores obtained on a measuring instrument are comparable to scores from an external criterion believed to measure the same concept.

Criterion variable The variable whose values are predicted from measurements of the predictor variable.

Cross-cultural comparisons Research studies that include culture as a major variable; studies that compare two or more diverse cultural groups.

Cross-sectional research design A survey research design in which data are collected to indicate characteristics of a sample or population at a particular moment in time.

Cross-tabulation table A simple table showing the joint frequency distribution of two or more nominal level variables.

Cultural encapsulation The assumption that differences between groups represent some deficit or pathology.

Culturally equivalent Similarity in the meaning of a construct between two cultures.

Cultural relativity The belief that human thought and action can be judged only from the perspective of the culture out of which they have grown.

Cut-and-paste method A method of analyzing qualitative data whereby the researcher cuts segments of the typed transcript and sorts these cuttings into relevant groupings; it can be done manually or with computer assistance.

Data The numbers, words, or scores, generated by positivist and interpretive research studies; the word data is plural.

Data analyses The process of turning data into information; the process of reviewing, summarizing, and organizing isolated facts (data) such that they formulate a meaningful response to a research question.

Data archive A place where many data sets are stored and from which data can be accessed.

Data coding Translating data from one language or format into another, usually to make it readable for a computer.

Data collection method Procedures specifying techniques to be employed, measuring instruments to be utilized, and activities to be conducted in implementing a positivist or interpretive research study.

Data set A collection of related data items, such as the answers given by respondents to all the questions in a survey.

Data source The provider of the data, whether it be primarythe original sourceor secondaryan intermediary between the research participant and the researcher analyzing the data.

Datum Singular of data.

Decision-making rule A statement that we use (in testing a hypothesis) to choose between the null hypothesis; indicates the range(s) of values of the observed statistic that leads to the rejection of the null hypothesis.

Deduction A conclusion about a specific case(s) based on the assumption that it shares a characteristic with an entire class of similar cases.

Deductive reasoning Forming a theory, making a deduction from the theory, and testing this deduction, or hypothesis, against reality; in research, applied to theory in order to arrive at a hypothesis that can be tested; a method of reasoning whereby a conclusion about specific cases is reached based on the assumption that they share characteristics with an entire class of similar cases.

Demand needs When needs are defined by only those individuals who indicate that they feel or perceive the need themselves.

Demographic data Vital and social facts that describe a sample or a population.

Demoralization Feelings of deprivation among control group members that may cause them to drop out of a research study; a threat to internal validity.

Dependability The soundness of both the steps taken in a qualitative data analysis and the conclusions reached.

Dependent events Events that influence the probability of occurrence of each other.

Dependent variable A variable that is dependent on, or caused by, another variable; an outcome variable, which is not manipulated directly but is measured to determine if the independent variable has had an effect.

Derived scores Raw scores of research participants, or groups, converted in such a way that meaningful comparisons with other individuals, or groups, are possible.

Descriptive research Research studies undertaken to increase precision in the definition of knowledge in a problem area where less is known than at the explanatory level; situated in the middle of the knowledge continuum.

Descriptive statistics Methods used for summarizing and describing data in a clear and precise manner.

Design bias Any effect that systematically distorts the outcome of a research study so that the studys results are not representative of the phenomenon under investigation.

Determinism A contention in positivist research studies that only an event that is true over time and place and that will occur independent of beliefs about it (a predetermined event) permits the generalization of a studys findings; one of the four main limitations of the positivist research approach.

Deterministic causation When a particular effect appears, the associated cause is always present; no other variables influence the relationship between cause and effect; the link between an independent variable that brings about the occurrence of the dependent variable every time.

Dichotomous variable A variable that can take on only one of two values.

Differential scale A questionnaire-type scale in which respondents are asked to consider questions representing different positions along a continuum and to select those with which they agree.

Differential selection A potential lack of equivalency among preformed groups of research participants; a threat to internal validity.

Diffusion of treatments Problems that may occur when experimental and control group members talk to each other about a research study; a threat to internal validity.

d index A measure of effect size in a meta-analysis.

Directional hypothesis See One-tailed hypotheses.

Directional test See One-tailed hypotheses.

Direct observation An obtrusive data collection method in which the focus is entirely on the behaviors of a group, or persons, being observed.

Direct observation notes These are the first level of field notes, usually chronologically organized, and they contain a detailed description of what was seen and heard; they may also include summary notes made after an interview.

Direct relationship A relationship between two variables such that high values of one variable are found with high values of the second variable, and vice versa.

Discriminant validity The degree to which a construct can be empirically differentiated, or discriminated from other constructs.

Divergent validity The extent to which a measuring instrument differs from other instruments that measure unrelated constructs.

Double-barreled question A question in a measuring instrument that contains two questions in one, usually joined by an and or an or.

Duration recording A method of data collection that includes direct observation of the target problem and recording of the length of time each occurrence lasts within a specified observation period.

Ecological fallacy An error of reasoning committed by coming to conclusions about individuals based only on data about groups.

Edge coding Adding a series of blank lines on the right side of the response category in a measuring instrument to aid in processing the data.

Effect size In meta-analysis, the most widely used measure of the dependent variable; the effect size statistic provides a measure of the magnitude of the relationship found between the variables of interest and allows for the computation of summary statistics that apply to the analysis of all the studies considered as a whole.

Empirical Knowledge derived from one of the ways of knowing.

Error of central tendency A measurement error due to the tendency of observers to rate respondents in the middle of a variables value range, rather than consistently too high or too low.

Error of measurement See Measurement error.

Ethical research project The systematic inquiry into a problem area in an effort to discover new knowledge or test existing ideas; the research study is conducted in accordance with professional standards.

Ethics in research Positivist and interpretive data that are collected and analyzed with careful attention to their accuracy, fidelity to logic, and respect for the feelings and rights of research participants; one of the four criteria for evaluating research problem areas and formulating research questions out of the problem areas.

Ethnicity A term that implies a common ancestry and cultural heritage and encompasses customs, values, beliefs, and behaviors.

Ethnocentricity Assumptions about normal behavior that are based on ones own cultural framework without taking cultural relativity into account; the failure to acknowledge alternative world views.

Ethnograph A computer software program that is designed for qualitative data analyses.

Ethnographic A form of content analysis used to document and explain the communication of meaning, as well as to verify theoretical relationships; any of several methods of describing social or cultural life based on direct, systematic observation, such as becoming a participant in a social system.

Ethnography The systematic study of human cultures and the similarities and dissimilarities between them.

Ethnomethodology Pioneered by Harold Garfinkel, this method of research focuses on the common-sense understanding of social life held by ordinary people (the ethos), usually as discovered through participant observation; often the observers own methods of making sense of the situation become the object of investigation.

Evaluation A form of appraisal using valid and reliable research methods; there are numerous types of evaluations geared to produce data that in turn produce information that helps in the decision-making process; data from evaluations are used to develop quality programs and services.

Evaluative research designs Case- and program-level research designs that apply various research designs and data collection methods to find out if an intervention (or treatment) worked at the case level and if the social work program worked at the program level.

Existing documents Physical records left over from the past.

Existing statistics Previously calculated numerical summaries of data that exist in the public domain.

Experience and intuition Learning what is true through personal past experiences and intuition; two of the ways of knowing.

Experiment A research study in which we have control over the levels of the independent variable and over the assignment of research participants, or objects, to different experimental conditions.

Experimental designs (1) Explanatory research designs or ideal experiments; (2) Case-level research designs that examine the question, Did the client system improve because of social work intervention?

Experimental group In an experimental research design, the group of research participants exposed to the manipulation of the independent variable; also referred to as a treatment group.

Explanatory research Ideal research studies undertaken to infer cause-effect and directional relationships in areas where a number of substantial research findings are already in place; situated at the top end of the knowledge continuum.

Exploratory research Research studies undertaken to gather data in areas of inquiry where very little is already known; situated at the lowest end of the knowledge continuum. See Nonexperimental design.

External evaluation An evaluation that is conducted by someone who does not have any connection with the program; usually an evaluation that is requested by the agencys funding sources; this type of evaluation complements an in-house evaluation.

External validity The extent to which the findings of a research study can be generalized outside the specific research situation.

Extraneous variables See Rival hypothesis.

Face validity The degree to which a measurement has self-evident meaning and measures what it appears to measure.

Feasibility One of the four criteria for evaluating research problem areas and formulating research questions out of the problem areas.

Feedback When data and information are returned to the persons who originally provided or collected them; used for informed decision making at the case and program levels; a basic principle underlying the design of evaluations.

Field notes A record, usually written, of events observed by a researcher; the notes are taken as the study proceeds, and later they are used for analyses.

Field research Research conducted in a real-life setting, not in a laboratory; the researcher neither creates nor manipulates anything within the study, but observes it.

Field-tested The pilot of an instrument or research method in conditions equivalent to those that will be encountered in the research study.

File drawer problem (1) In literature searches or reviews, the difficulty in locating studies that have not been published or are not easily retrievable; (2) In meta-analyses, errors in effect size due to reliance on published articles showing statistical significance.

Firsthand data Data obtained from people who directly experience the problem being studied.

First-level coding A process of identifying meaning units in a transcript, organizing the meaning units into categories, and assigning names to the categories.

Flexibility The degree to which the design and procedures of a research study can be changed to adapt to contextual demands of the research setting.

Focus group interview A group of people brought together to talk about their lives and experiences in free-flowing, open-ended discussions that usually focus on a single topic.

Formative evaluation A type of evaluation that focuses on obtaining data that are helpful in planning the program and in improving its implementation and performance.

Frequency recording A method of data collection by direct observations in which each occurrence of the target problem is recorded during a specified observation period.

Fugitive data Informal information found outside regular publishing channels.

Gaining access A term used in interpretive research to describe the process of engagement and relationship development between the researcher and the research participants.

Generalizable explanation evaluation model An evaluation model whose proponents believe that many solutions are possible for any one social problem and that the effects of programs will differ under different conditions.

Generalizing results Extending or applying the findings of a research study to individuals or situations not directly involved in the original research study; the ability to extend or apply the findings of a research study to subjects or situations that were not directly investigated.

Goal Attainment Scale (GAS) A modified measurement scale used to evaluate case or program outcomes.

Government documents Printed documents issued by local, state, and federal governments; such documents include reports of legislative committee hearings and investigations, studies commissioned by legislative commissions and executive agencies, statistical compilations such as the census, the regular and special reports of executive agencies, and much more.

Grand tour questions Queries in which research participants are asked to provide wide-ranging background information; mainly used in interpretive research studies.

Graphic rating scale A rating scale that describes an attribute on a continuum from one extreme to the other, with points of the continuum ordered in equal intervals and then assigned values.

Grounded theory A final outcome of the interpretive research process that is reached when the insights are grounded on observations and the conclusions seem to be firm.

Group evaluation designs Evaluation designs that are conducted with groups of cases for the purpose of assessing to what degree program objectives have been achieved.

Group research designs Research designs conducted with two or more groups of cases, or research participants, for the purpose of answering research questions or testing hypotheses.

Halo effect A measurement error due to the tendency of an observer to be influenced by a favorable trait(s) of a research participant(s).

Hawthorne effect Effects on research participants behaviors or attitudes attributable to their knowledge that they are taking part in a research study; a reactive effect; a threat to external validity.

Heterogeneity of respondents The extent to which a research participant differs from other research participants.

Heuristic A theory used to stimulate creative thought and scientific activity.

Historical research The process by which we study the past; a method of inquiry that attempts to explain past events based on surviving artifacts.

History in research design The possibility that events not accounted for in a research design may alter the second and subsequent measurements of the dependent variable; a threat to internal validity.

Homogeneity of respondents The extent to which a research participant is similar to other research participants.

Hypothesis A theory-based prediction of the expected results of a research study; a tentative explanation that a relationship between or among variables exists.

Hypothetico-deductive method A hypothesis-testing approach that a hypothesis is derived on the deductions based from a theory.

Ideographic research Research studies that focus on unique individuals or situations.

Implementation of a program The action of carrying out a program in the way that it was designed.

Independent variable A variable that is not dependent on another variable but is believed to cause or determine changes in the dependent variable; an antecedent variable that is directly manipulated in order to assess its effect on the dependent variable.

Index A group of individual measures that, when combined, are meant to indicate some more general characteristic.

Indigenous observers People who are naturally a part of the research participants environment and who perform the data collection function; includes relevant others (e.g., family members, peers) and collaterals (e.g., social workers, staff members).

Indirect measures A substitute variable, or a collection of representative variables, used when there is no direct measurement of the variable of interest; also called a proxy variable.

Individualism A way of living that stresses independence, personal rather than group objectives, competition, and power in relationships; achievement measured through success of the individual as opposed to the group.

Individual synthesis Analysis of published studies related to the subject under study.

Inductive reasoning Building on specific observations of events, things, or processes to make inferences or more general statements; in research studies, applied to data collection and research results to make generalizations to see if they fit a theory; a method of reasoning whereby a conclusion is reached by building on specific observations of events, things, or processes to make inferences or more general statements.

Inferential statistics Statistical methods that make it possible to draw tentative conclusions about the population based on observations of a sample selected from that population and, furthermore, to make a probability statement about those conclusions to aid in their evaluation.

Information anxiety A feeling attributable to a lack of understanding of information, being overwhelmed by the amount of information to be accessed and understood, or not knowing if certain information exists.

Informed consent Signed statements obtained from research participants prior to the initiation of the research study to inform them what their participation entails and that they are free to decline participation.

In-house evaluation An evaluation that is conducted by someone who works within a program; usually an evaluation for the purpose of promoting better client services; also known as an internal evaluation; this type of evaluation complements an external evaluation.

Institutional review boards (IRBs) Boards set up by institutions in order to protect research participants and to ensure that ethical issues are recognized and responded to in the a studys research design.

Instrumentation Weaknesses of a measuring instrument, such as invalidity, unreliability, improper administrations, or mechanical breakdowns; a threat to internal validity.

Integration Combining evaluation and day-to-day practice activities to develop a complete approach to client service delivery; a basic principle underlying the design of evaluations.

Interaction effects Effects produced by the combination of two or more threats to internal validity.

Internal consistency The extent to which the scores on two comparable halves of the same measuring instrument are similar; inter-item consistency.

Internal validity The extent to which it can be demonstrated that the independent variable within a research study is the only cause of change in the dependent variable; overall soundness of the experimental procedures and measuring instruments.

Interpretive research approach Research studies that focus on the facts of nature as they occur under natural conditions and emphasize qualitative description and generalization; a process of discovery sensitive to holistic and ecological issues; a research approach that is complementary to the positivist research approach.

Interobserver reliability The stability or consistency of observations made by two or more observers at one point in time.

Interpretive notes Notes on the researchers interpretations of events that are kept separate from the record of the facts noted as direct observations.

Interquartile range A number that measures the variability of a data set; the distance between the 75th and 25th percentiles.

Interrater reliability The degree to which two or more independent observers, coders, or judges produce consistent results.

Interrupted time-series design An explanatory research design in which there is only one group of research participants and the dependent variable is measured repeatedly before and after treatment; used in case- and program-evaluation designs.

Interval level of measurement The level of measurement with an arbitrarily chosen zero point that classifies its values on an equally spaced continuum.

Interval recording A method of data collection that involves a continuous direct observation of an individual during specified observation periods divided into equal time intervals.

Intervening variable See Rival hypothesis.

Interview data Isolated facts that are gathered when research participants respond to carefully constructed research questions; data, which are in the form of words, are recorded by transcription.

Interviewing A conversation with a purpose.

Interview schedule A measuring instrument used to collect data in face-to-face and telephone interviews.

Intraobserver reliability The stability of observations made by a single observer at several points in time.

Intrusion into lives of research participants The understanding that specific data collection methods can have negative consequences for research participants; a criterion for selecting a data collection method.

Itemized rating scales A measuring instrument that presents a series of statements that respondents or observers rank in different positions on a specific attribute.

Journal A written record of the process of an interpretive research study. Journal entries are made on an ongoing basis throughout the study and include study procedures as well as the researchers reactions to emerging issues and concerns during the data analysis process.

Key informants A subpopulation of research participants who seem to know much more about the situation than other research participants.

Knowledge base A body of knowledge and skills specific to a certain discipline.

Knowledge creator and disseminator A social work role in which the social worker actually carries out and disseminates the results of a positivist and/or interpretive research study to generate knowledge for our profession.

Knowledge level continuum The range of knowledge levels, from exploratory to descriptive to explanatory, at which research studies can be conducted.

Latent content In a content analysis, the true meaning, depth, or intensity of a variable, or concept, under study.

Levels of measurement The degree to which characteristics of a data set can be modeled mathematically; the higher the level of measurement, the more statistical methods that are applicable.

Limited review An existing literature synthesis that summarizes in narrative form the findings and implications of a few research studies.

Literature review See Literature search and Review of the literature.

Literature search In a meta-analysis, scanning books and journals for basic, up-to-date research articles on studies relevant to a research question or hypothesis; sufficiently thorough to maximize the chance of including all relevant sources. See Review of the literature.

Logical consistency The requirement that all the steps within a positivist research study must be logically related to one another.

Logical positivism A philosophy of science holding that the scientific method of inquiry is the only source of certain knowledge; in research, focuses on testing hypotheses deduced from theory.

Logistics In evaluation, refers to getting research participants to do what they are supposed to do, getting research instruments distributed and returned; in general, the activities that ensure that procedural tasks of a research or evaluation study are carried out.

Longitudinal case study An exploratory research design in which there is only one group of research participants and the dependent variable is measured more than once.

Longitudinal design A survey research design in which a measuring instrument(s) is administered to a sample of research participants repeatedly over time; used to detect dynamic processes such as opinion change.

Magnitude recording A direct-observation method of soliciting and recording data on amount, level, or degree of the target problem during each occurrence.

Management information system (MIS) System in which computer technology is used to process, store, retrieve, and analyze data collected routinely in such processes as social service delivery.

Manifest content Content of a communication that is obvious and clearly evident.

Manipulable solution evaluation model An evaluation model whose proponents believe that the greatest priority is to serve the public interest, not the interests of stakeholders, who have vested interests in the program being evaluated; closely resembles an outcome evaluation.

Matching A random assignment technique that assigns research participants to two or more groups so that the experimental and control groups are approximately equivalent in pretest scores or other characteristics, or so that all differences except the experimental condition are eliminated.

Maturation Unplanned change in research participants due to mental, physical, or other processes operating over time; a threat to internal validity.

Meaning units In a qualitative data analysis, a discrete segment of a transcript that can stand alone as a single idea; can consist of a single word, a partial or complete sentence, a paragraph, or more; used as the basic building blocks for developing categories.

Measurement The assignment of labels or numerals to the properties or attributes of observations, events, or objects according to specific rules.

Measurement error Any variation in measurement that cannot be attributed to the variable being measured; variability in responses produced by individual differences and other extraneous variables.

Measuring instrument Any instrument used to measure a variable(s).

Media myths The content of television shows, movies, and newspaper and magazine articles; one of the six ways of knowing.

Member checking A process of obtaining feedback and comments from research participants on interpretations and conclusions made from the qualitative data they provided; asking research participants to confirm or refute the conclusions made.

Meta-analysis A research method in which mathematical procedures are applied to the positivist findings of studies located in a literature search to produce new summary statistics and to describe the findings for a meta-analysis.

Methodology The procedures and rules that detail how a single research study is conducted.

Micro-level data Data derived from individual units of analysis, whether these data sources are individuals, families, corporations, etc.; for example, age and years of formal schooling are two variables requiring micro-level data.

Missing data Data not available for a research participant about whom other data are available, such as when a respondent fails to answer one of the questions in a survey.

Missing links When two categories or themes seem to be related, but not directly so, it may be that a third variable connects the two.

Mixed research model A model combining aspects of interpretive and positivist research approaches within all (or many) of the methodological steps contained within a single research study.

Monitoring approach to evaluation Evaluation that aims to provide ongoing feedback so that a program can be improved while it is still underway; it contributes to the continuous development and improvement of a human service program; this approach complements the project approach to evaluation.

Mortality Loss of research participants through normal attrition over time in an experimental design that requires retesting; a threat to internal validity.

Multicultural research Representation of diverse cultural factors in the subjects of study; such diversity variables may include religion, race, ethnicity, language preference, gender, etc.

Multigroup posttest-only design An exploratory research design in which there is more than one group of research participants and the dependent variable is measured only once for each group.

Multiple-baseline design A case-level evaluation design with more than one baseline period and intervention phase, which allows the causal inferences regarding the relationship between a treatment intervention and its effect on clients target problems and which helps control for extraneous variables. See Interrupted time-series design.

Multiple-group design An experimental research design with one control group and several experimental groups.

Multiple-treatment interference Effects of the results of a first treatment on the results of second and subsequent treatments; a threat to external validity.

Multistage probability sampling Probability sampling procedures used when a comprehensive list of the population does not exist and it is not possible to construct one.

Multivariate (1) A relationship involving two or more variables; (2) A hypothesis stating an assertion about two or more variables and how they relate to one another.

Multivariate analysis A statistical analysis of the relationship among three or more variables.

Narrowband measuring instrument Measuring instruments that focus on a single, or a few, variables.

Nationality A term that refers to country of origin.

Naturalist A person who studies the facts of nature as they occur under natural conditions.

Needs assessment Program-level evaluation activities that aim to assess the feasibility for establishing or continuing a particular social service program; an evaluation that aims to assess the need for a human service by verifying that a social problem exists within a specific client population to an extent that warrants services.

Negative case sampling Purposefully selecting research participants based on the fact that they have different characteristics than previous cases.

Nominal level of measurement The level of measurement that classifies variables by assigning names or categories that are mutually exclusive and exhaustive.

Nondirectional test See Two-tailed hypotheses.

Nonexperimental design A research design at the exploratory, or lowest, level of the knowledge continuum; also called preexperimental.

Nonoccurrence data In the structured-observation method of data collection, a recording of only those time intervals in which the target problem did not occur.

Nonparametric tests Refers to statistical tests of hypotheses about population probability distributions, but not about specific parameters of the distributions.

Nonprobability sampling Sampling procedures in which all of the persons, events, or objects in the sampling frame have an unknown, and usually different, probability of being included in a sample.

Nonreactive Methods of research that do not allow the research participants to know that they are being studied; thus, they do not alter their responses for the benefit of the researcher.

Nonresponse The rate of nonresponse in survey research is calculated by dividing the total number of respondents by the total number in the sample, minus any units verified as ineligible.

Nonsampling errors Errors in a research studys results that are not due to the sampling procedures.

Norm In measurement, an average or set group standard of achievement that can be used to interpret individual scores; normative data describing statistical properties of a measuring instrument such as means and standard deviations.

Normalization group The population sample to which a measuring instrument under development is administered in order to establish norms; also called the norm group.

Normative needs When needs are defined by comparing the objective living conditions of a target population with what societyor, at least, that segment of society concerned with helping the target populationdeems acceptable or desirable from a humanitarian standpoint.

Null hypothesis A statement concerning one or more parameters that is subjected to a statistical test; a statement that there is no relationship between the two variables of interest.

Numbers The basic data unit of analysis used in positivist research studies.

Objectivity A research stance in which a study is carried out and its data are examined and interpreted without distortion by personal feelings or biases.

Observer One of four roles on a continuum of participation in participant observation research; the level of involvement of the observer participant is lower than of the complete participant and higher than of the participant observer.

Obtrusive data collection methods Direct data collection methods that can influence the variables under study or the responses of research participants; data collection methods that produce reactive effects.

Occurrence data In the structured-observation method of data collection, a recording of the first occurrence of the target problem during each time interval.

One-group posttest-only design An exploratory research design in which the dependent variable is measured only once.

One-group pretest-posttest design A descriptive research design in which the dependent variable is measured twicebefore and after treatment.

One-stage probability sampling Probability sampling procedures in which the selection of a sample that is drawn from a specific population is completed in a single process.

One-tailed hypotheses Statements that predict specific relationships between independent and dependent variables.

On-line sources Computerized literary retrieval systems that provide printouts of indexes and abstracts.

Open-ended questions Unstructured questions in which the response categories are not specified or detailed.

Operational definition Explicit specification of a variable in such a way that its measurement is possible.

Operationalization The process of developing operational definitions of the variables that are contained within the concepts of a positivist and/or interpretive research study.

Ordinal level of measurement The level of measurement that classifies variables by rank-ordering them from high to low or from most to least.

Outcome The effect of the manipulation of the independent variable on the dependent variable; the end product of a treatment intervention.

Outcome measure The criterion or basis for measuring effects of the independent variable or change in the dependent variable.

Outcome-oriented case study A type of case study that investigates whether client outcomes were in fact achieved.

Outside observers Trained observers who are not a part of the research participants environment and who are brought in to record data.

Paired observations An observation on two variables, where the intent is to examine the relationship between them.

Panel research study A longitudinal survey design in which the same group of research participants (the panel) is followed over time by surveying them on successive occasions.

Parametric tests Statistical methods for estimating parameters or testing hypotheses about population parameters.

Participant observation An obtrusive data collection method in which the researcher, or the observer, participates in the life of those being observed; both an obtrusive data collection method and a research approach, this method is characterized by the one doing the study undertaking roles that involve establishing and maintaining ongoing relationships with research participants who are often in the field settings, and observing and participating with the research participants over time.

Participant observer The participant observer is one of four roles on a continuum of participation in participant observation research; the level of involvement of the participant observer is higher than of the complete observer and lower than of the observer participant.

Permanent product recording A method of data collection in which the occurrence of the target problem is determined by observing the permanent product or record of the target problem.

Pilot study See Pretest (2).

Population An entire set, or universe, of people, objects, or events of concern to a research study, from which a sample is drawn.

Positivism See positivist research approach.

Positivist research approach A research approach to discover relationships and facts that are generalizable; research that is independent of subjective beliefs, feelings, wishes, and values; a research approach that is complementary to the interpretive research approach.

Posttest Measurement of the dependent variable after the introduction of the independent variable.

Potential for testing One of the four criteria for evaluating research hypotheses.

Practitioner/researcher A social worker who guides practice through the use of research findings; collects data throughout an intervention using research methods, skills, and tools; disseminates practice findings.

Pragmatists Researchers who believe that both interpretive and positivist research approaches can be integrated in a single research study.

Predictive validity A form of criterion validity that is concerned with the ability of a measuring instrument to predict future performance or status on the basis of present performance or status.

Predictor variable The variable that, it is believed, allows us to improve our ability to predict values of the criterion variable.

Preexposure Tasks to be carried out in advance of a research study to sensitize the researcher to the culture of interest; these tasks may include participation in cultural experiences, intercultural sharing, case studies, ethnic literature reviews, value statement exercises, etc.

Preliminary plan for data analysis A strategy for analyzing qualitative data that is outlined in the beginning stages of an interpretive research study; the plan has two general steps: (1) previewing the data, and (2) outlining what to record in the researchers journal.

Presentism Applying current thinking and concepts to interpretations of past events or intentions.

Pretest (1) Measurement of the dependent variable prior to the introduction of the independent variable; (2) Administration of a measuring instrument to a group of people who will not be included in the study to determine difficulties the research participants may have in answering questions and the general impression given by the instrument; also called a pilot study.

Pretest-treatment interaction Effects that a pretest has on the responses of research participants to the introduction of the independent variable or the experimental treatment; a threat to external validity.

Previous research Research studies that have already been completed and published; they provide information about data collection methods used to investigate research questions that are similar to our own; a criterion for selecting a data collection method.

Primary data Data in its original form, as collected from the research participants; a primary data source is one that puts as few intermediaries as possible between the production and the study of the data.

Primary language The preferred language of the research participants.

Primary reference source A report of a research study by the person who conducted the study; usually an article in a professional journal.

Probability sampling Sampling procedures in which every member of the designated population has a known probability of being selected for the sample.

Problem area In social work research, a general expressed difficulty about which something researchable is unknown; not to be confused with research question.

Problem-solving process A generic method with specified phases for solving problems; also described as the scientific method.

Process-oriented case study A type of case study that illuminates the micro-steps of intervention that lead to client outcomes; describes how programs and interventions work and gives insight into the black box of intervention.

Professional standards Rules for making judgments about evaluation activity that are established by a group of persons who have advanced education and usually have the same occupation.

Program An organized set of political, administrative, and clinical activities that function to fulfill some social purpose.

Program development The constant effort to improve program services to better achieve outcomes; a basic principle underlying the design of evaluations.

Program efficiency Assessment of a programs outcome in relation to the costs of obtaining the outcome.

Program evaluation A form of appraisal, using valid and reliable research methods, that examines the processes or outcomes of an organization that exists to fulfill some social purpose.

Program goal A statement defining the intent of a program that cannot be directly evaluated; it can, however, be evaluated indirectly by the programs objectives, which are derived from the program goal; not to be confused with program objectives.

Program-level evaluation A form of appraisal that monitors change for groups of clients and organizational performance.

Program objectives A statement that clearly and exactly specifies the expected change, or intended result, for individuals receiving program services; qualities of well-chosen objectives are meaningfulness, specificity, measurability, and directionality; not to be confused with program goal.

Program participation The philosophy and structure of a program that will support or supplant the successful implementation of a research study within an existing social service program; a criterion for selecting a data collection method.

Program process The coordination of administrative and clinical activities that are designed to achieve a programs goal.

Program results A report on how effective a program is at meeting its stated objectives.

Project approach to evaluation Evaluation that aims to assess a completed or finished program; this approach complements the monitoring approach.

Proxy An indirect measure of a variable that a researcher wants to study; it is often used when the variable of inquiry is difficult to measure or observe directly.

Pure research approach A search for theoretical results that can be utilized to develop theory and expand our professions knowledge bases; complementary to the applied research approach.

Purists Researchers who believe that interpretive and positivist research approaches should never be mixed.

Purpose statement A declaration of words that clearly describes a research studys intent.

Purposive sampling A nonprobability sampling procedure in which research participants with particular characteristics are purposely selected for inclusion in a research sample; also known as judgmental or theoretical sampling.

Qualitative data Data that measure a quality or kind; when referring to variables, qualitative is another term for categorical or nominal variable values; when speaking of kinds of research, qualitative refers to studies of subjects that are hard to quantify; interpretive research produces descriptive data based on spoken or written words and observable behaviors.

Quantification In measurement, the reduction of data to numerical form in order to analyze them by way of mathematical or statistical techniques.

Quantitative data Data that measure a quantity or amount.

Quasi-experiment A research design at the descriptive level of the knowledge continuum that resembles an ideal experiment but does not allow for random selection or assignment of research participants to groups and often does not control for rival hypotheses.

Questionnaire-type scale A type of measuring instrument in which multiple responses are usually combined to form a single overall score for a respondent.

Quota sampling A nonprobability sampling procedure in which the relevant characteristics of the sample are identified, the proportion of these characteristics in the population is determined, and research participants are selected from each category until the predetermined proportion (quota) has been achieved.

Race A variable based on physical attributes that can be subdivided into the Caucasoid, Negroid, and Mongoloid races.

Random assignment The process of assigning individuals to experimental or control groups so that the groups are equivalent; also referred to as randomization.

Random error Variable error in measurement; error due to unknown or uncontrolled factors that affect the variable being measured and the process of measurement in an inconsistent fashion.

Randomized cross-sectional survey design A descriptive research design in which there is only one group, the dependent variable is measured only once, the research participants are randomly selected from the population, and there is no independent variable.

Randomized longitudinal survey design A descriptive research design in which there is only one group, the dependent variable is measured more than once, and research participants are randomly selected from the population before each treatment.

Randomized one-group posttest-only design A descriptive research design in which there is only one group, the dependent variable is measured only once, and research participants are randomly selected from the population.

Randomized posttest-only control group design An explanatory research design in which there are two or more randomly assigned groups, the control group does not receive treatment, and the experimental groups receive different treatments.

Random numbers table A computer-generated or published table of numbers in which each number has an equal chance of appearing in each position in the table.

Random sampling An unbiased selection process conducted so that all members of a population have an equal chance of being selected to participate in a research study.

Rank-order scale A comparative rating scale in which the rater is asked to rank specific individuals in relation to one another on some characteristic.

Rating scale A type of measuring instrument in which responses are rated on a continuum or in an ordered set of categories, with numerical values assigned to each point or category.

Ratio level of measurement The level of measurement that has a nonarbitrary, fixed zero point and classifies the values of a variable on an equally spaced continuum.

Raw scores Scores derived from administration of a measuring instrument to research participants or groups.

Reactive effect (1) An effect on outcome measures due to the research participants awareness that they are being observed or interviewed; a threat to external and internal validity; (2) Alteration of the variables being measured or the respondents performance on the measuring instrument due to administration of the instrument.

Reactivity The belief that things being observed or measured are affected by the fact that they are being observed or measured; one of the four main limitations of the positivist research approach.

Reassessment A step in a qualitative data analysis in which the researcher interrupts the data analysis process to reaffirm the rules used to decide which meaning units are placed within different categories.

Recoding Developing and applying new variable value labels to a variable that has previously been coded; usually, recoding is done to make variables from one or more data sets comparable.

Reductionism In the positivist research approach, the operationalization of concepts by reducing them to common measurable variables; one of the four main limitations of the positivist research approach.

Relevancy One of the four criteria for evaluating research problem areas and formulating research questions out of the problem areas.

Reliability (1) The degree of accuracy, precision, or consistency in results of a measuring instrument, including the ability to produce the same results when the same variable is measured more than once or repeated applications of the same test on the same individual produce the same measurement; (2) The degree to which individual differences on scores or in data are due either to true differences or to errors in measurement.

Replication Repetition of the same research procedures by a second researcher for the purpose of determining if earlier results can be confirmed.

Researchability The extent to which a research problem is in fact researchable and the problem can be resolved through the consideration of data derived from a research study; one of the four criteria for evaluating research problem areas and formulating research questions out of the problem areas.

Research attitude A way that we view the world. It is an attitude that highly values craftsmanship, with pride in creativity, high-quality standards, and hard work.

Research consumer A social work role reflecting the ethical obligation to base interventions on the most up-to-date research knowledge available.

Research design The entire plan of a positivist and/or interpretive research study from problem conceptualization to the dissemination of findings.

Researcher bias The tendency of researchers to find results they expect to find; a threat to external validity.

Research hypothesis A statement about a studys research question that predicts the existence of a particular relationship between the independent and dependent variables; can be used in both the positivist and interpretive approaches to research.

Research method The use of positivist and interpretive research approaches to find out what is true; one of the ways of knowing.

Research participants People utilized in research studies; also called subjects or cases.

Research question A specific research question that is formulated directly out of the general research problem area; answered by the interpretive and/or positivist research approach; not to be confused with problem area.

Resources The costs associated with collecting data in any given research study; includes materials and supplies, equipment rental, transportation, training staff, and staff time; a criterion for selecting a data collection method.

Response categories Possible responses assigned to each question in a standardized measuring instrument, with a lower value generally indicating a low level of the variable being measured and a larger value indicating a higher level.

Response rate The total number of responses obtained from potential research participants to a measuring instrument divided by the total number of responses requested, usually expressed in the form of a percentage.

Response set Personal style; the tendency of research participants to respond to a measuring instrument in a particular way, regardless of the questions asked, or the tendency of observers or interviewers to react in certain ways; a source of constant error.

Review of the literature (1) A search of the professional literature to provide background knowledge of what has already been examined or tested in a specific problem area; (2) Use of any information source, such as a computerized data base, to locate existing data or information on a research problem, question, or hypothesis.

Rival hypothesis A hypothesis that is a plausible alternative to the research hypothesis and might explain the results as well or better; a hypothesis involving extraneous or intervening variables other than the independent variable in the research hypothesis; also referred to as an alternative hypothesis.

Rules of correspondence A characteristic of measurement stipulating that numerals or symbols are assigned to properties of individuals, objects, or events according to specified rules.

Sample A subset of a population of individuals, objects, or events chosen to participate in or to be considered in a research study.

Sampling error (1) The degree of difference that can be expected between the sample and the population from which it was drawn; (2) A mistake in a research studys results that is due to sampling procedures.

Sampling frame A listing of units (people, objects, or events) in a population from which a sample is drawn.

Sampling plan A method of selecting members of a population for inclusion in a research study, using procedures that make it possible to draw inferences about the population from the sample statistics.

Sampling theory The logic of using methods to ensure that a sample and a population are similar in all relevant characteristics.

Scale A measuring instrument composed of several items that are logically or empirically structured to measure a construct.

Scattergram A graphic representation of the relationship between two interval- or ratio-level variables.

Science Knowledge that has been obtained and tested through use of positivist and interpretive research studies.

Scientific community A group that shares the same general norms for both research activity and acceptance of scientific findings and explanations.

Scientific determinism See Determinism.

Scientific method A generic method with specified steps for solving problems; the principles and procedures used in the systematic pursuit of knowledge.

Scope of a study The extent to which a problem area is covered in a single research study; a criterion for selecting a data collection method.

Score A numerical value assigned to an observation; also called data.

Search statement A preliminary search statement developed by the researcher prior to a literature search and which contains terms that can be combined to elicit specific data.

Secondary analysis An unobtrusive data collection method in which available data that predate the formulation of a research study are used to answer the research question or test the hypothesis.

Secondary data Data that predate the formulation of the research study and which are used to answer the research question or test the hypothesis.

Secondary data sources A data source that provides nonoriginal, secondhand data.

Secondary reference source A source related to a primary source or sources, such as a critique of a particular source item or a literature review, bibliography, or commentary on several items.

Secondhand data Data obtained from people who are indirectly connected to the problem being studied.

Selection-treatment interaction The relationship between the manner of selecting research participants and their response to the independent variable; a threat to external validity.

Self-anchored scales A rating scale in which research participants rate themselves on a continuum of values, according to their own referents for each point.

Self-disclosure Shared communication about oneself, including ones behaviors, beliefs, and attitudes.

Semantic differential scale A modified measurement scale in which research participants rate their perceptions of the variable under study along three dimensionsevaluation, potency, and activity.

Sequential triangulation When two distinct and separate phases of a research study are conducted and the results of the first phase are considered essential for planning the second phase; research questions in Phase 1 are answered before research questions in Phase 2 are formulated.

Service recipients People who use human servicesindividuals, couples, families, groups, organizations, and communities; also known as clients or consumers; a stakeholder group in evaluation.

Simple random sampling A one-stage probability sampling procedure in which members of a population are selected one at a time, without a chance of being selected again, until the desired sample size is obtained.

Simultaneous triangulation When the results of a positivist and interpretive research question are answered at the same time; results to the interpretive research questions, for example, are reported separately and do not necessarily relate to, or confirm, the results from the positivist phase.

Situationalists Researchers who assert that certain research approaches (interpretive or positivist) are appropriate for specific situations.

Situation-specific variable A variable that may be observable only in certain environments and under certain circumstances, or with particular people.

Size of a study The number of people, places, or systems that are included in a single research study; a criterion for selecting a data collection method.

Snowball sampling A nonprobability sampling procedure in which individuals selected for inclusion in a sample are asked to identify other individuals from the population who might be included; useful to locate people with divergent points of view.

Social desirability (1) A response set in which research participants tend to answer questions in a way that they perceive as giving favorable impressions of themselves; (2) The inclination of data providers to report data that present a socially desirable impression of themselves or their reference groups. Also referred to as impression management.

Socially acceptable response Bias in an answer that comes from research participants trying to answer questions as they think a good person should, rather than in a way that reveals what they actually believe or feel.

Social work research Scientific inquiry in which interpretive and positivist research approaches are used to answer research questions and create new, generally applicable knowledge in the field of social work.

Socioeconomic variables Any one of several measures of social rank, usually including income, education, and occupational prestige; abbreviated SES.

Solomon four-group design An explanatory research design with four randomly assigned groups, two experimental and two control; the dependent variable is measured before and after treatment for one experimental and one control group, but only after treatment for the other two groups, and only experimental groups receive the treatment.

Specificity One of the four criteria for evaluating research hypotheses.

Split-half method A method for establishing the reliability of a measuring instrument by dividing it into comparable halves and comparing the scores between the two halves.

Spot-check recording A method of data collection that involves direct observation of the target problem at specified intervals rather than on a continuous basis.

Stakeholder A person or group of people having a direct or indirect interest in the results of an evaluation.

Stakeholder service evaluation model Proponents of this evaluation model believe that program evaluations will be more likely to be utilized, and thus have a greater impact on social problems, when they are tailored to the needs of stakeholders; in this model, the purpose of program evaluation is not to generalize findings to other sites, but rather to restrict the evaluation effort to a particular program.

Standardized measuring instrument A professionally developed measuring instrument that provides for uniform administration and scoring and generates normative data against which later results can be evaluated.

Statistics The branch of mathematics concerned with the collection and analysis of data using statistical techniques.

Stratified random sampling A one-stage probability sampling procedure in which a population is divided into two or more strata to be sampled separately, using simple random or systematic random sampling techniques.

Structured interview schedule A complete list of questions to be asked and spaces for recording the answers; the interview schedule is used by interviewers when questioning respondents.

Structured observation A data collection method in which people are observed in their natural environments using specified methods and measurement procedures. See Direct observation.

Subscale A component of a scale that measures some part or aspect of a major construct; also composed of several items that are logically or empirically structured.

Summated scale A questionnaire-type scale in which research participants are asked to indicate the degree of their agreement or disagreement with a series of questions.

Summative evaluation A type of evaluation that examines the ultimate success of a program and assists with decisions about whether a program should be continued or chosen in the first place among alternative program options.

Survey research A data collection method that uses survey-type data collection measuring instruments to obtain opinions or answers from a population or sample of research participants in order to describe or study them as a group.

Synthesis Undertaking the search for meaning in our sources of information at every step of the research process; combining parts such as data, concepts, and theories to arrive at a higher level of understanding.

Systematic To arrange the steps of a research study in a methodical way.

Systematic random sampling A one-stage probability sampling procedure in which every person at a designated interval in a specific population is selected to be included in a research studys sample.

Systematic error Measurement error that is consistent, not random.

Target population The group about which a researcher wants to draw conclusions; another term for a population about which one aims to make inferences.

Target problem (1) In case-level evaluation designs, the problems social workers seek to solve for their clients; (2) A measurable behavior, feeling, or cognition that is either a problem in itself or symptomatic of some other problem.

Temporal research design A research study that includes time as a major variable; the purpose of this design is to investigate change in the distribution of a variable or in relationships among variables over time; there are three types of temporal research designs: cohort, panel, and trend.

Temporal stability Consistency of responses to a measuring instrument over time; reliability of an instrument across forms and across administrations.

Testing effect The effect that taking a pretest might have on posttest scores; a threat to internal validity.

Test-retest reliability Reliability of a measuring instrument established through repeated administration to the same group of individuals.

Thematic notes In observational research, thematic notes are a record of emerging ideas, hypotheses, theories, and conjectures; thematic notes provide a place for the researcher to speculate and identify themes, make linkages between ideas and events, and articulate thoughts as they emerge in the field setting.

Theme In a qualitative data analysis, a concept or idea that describes a single category or a grouping of categories; an abstract interpretation of qualitative data.

Theoretical framework A frame of reference that serves to guide a research study and is developed from theories, findings from a variety of other studies, and the researchers personal experiences.

Theoretical sampling See Purposive sampling.

Theory A reasoned set of propositions, derived from and supported by established data, which serves to explain a group of phenomena; a conjectural explanation that may, or may not, be supported by data generated from interpretive and positivist research studies.

Time orientation An important cultural factor that considers whether one is future-, present-, or past-oriented; for instance, individuals who are present-oriented would not be as preoccupied with advance planning as those who are future-oriented.

Time-series design See Interrupted time-series design.

Tradition Traditional cultural beliefs that we accept without question as true; one of the ways of knowing.

Transcript A written, printed, or typed copy of interview data or any other written material that have been gathered for an interpretive research study.

Transition statements Sentences used to indicate a change in direction or focus of questions in a measuring instrument.

Treatment group See Experimental group.

Trend study A longitudinal study design in which data from surveys carried out at periodic intervals on samples drawn from a particular population are used to reveal trends over time.

Triangulation The idea of combining different research methods in all steps associated with a single research study; assumes that any bias inherent in one particular method will be neutralized when used in conjunction with other research methods; seeks convergence of a studys results; using more than one research method and source of data to study the same phenomena and to enhance validity; there are several types of triangulation, but the essence of the term is that multiple perspectives are compared; it can involve multiple data sources or multiple data analyzers; the hope is that the different perspectives will confirm each other, adding weight to the credibility and dependability of qualitative data analysis.

Triangulation of analysts Using multiple data analyzers to code a single segment of transcript and comparing the amount of agreement between analyzers; a method used to verify coding of qualitative data.

Two-phase research model A model combining interpretive and positivist research approaches in a single study where each approach is conducted as a separate and distinct phase of the study.

Two-tailed hypotheses Statements that do not predict specific relationships between independent and dependent variables.

Unit of analysis A specific research participant (person, object, or event) or the sample or population relevant to the research question; the persons or things being studied; units of analysis in research are often persons, but may be groups, political parties, newspaper editorials, unions, hospitals, schools, etc.; a particular unit of analysis from which data are gathered is called a case.

Univariate A hypothesis or research design involving a single variable.

Universe See Population.

Unobtrusive methods Data collection methods that do not influence the variable under study or the responses of research participants; methods that avoid reactive effects.

Unstructured interviews A series of questions that allow flexibility for both the research participant and the interviewer to make changes during the process.

Validity (1) The extent to which a measuring instrument measures the variable it is supposed to measure and measures it accurately; (2) The degree to which an instrument is able to do what it is intended to do, in terms of both experimental procedures and measuring instruments (internal validity) and generalizability of results (external validity); (3) The degree to which scores on a measuring instrument correlate with measures of performance on some other criterion.

Variable A concept with characteristics that can take on different values.

Verbatim recording Recording interview data word-for-word and including significant gestures, pauses, and expressions of persons in the interview.

Wideband measuring instrument An instrument that measures more than one variable.

Within-methods research approach Triangulation by using different research methods available in either the interpretive or the positivist research approaches in a single research study.

Words The basic data unit of analysis used in interpretive research studies.

Worker cooperation The actions and attitudes of program personnel when carrying out a research study within an existing program; a criterion for selecting a data collection method.

Working hypothesis An assertion about a relationship between two or more variables that may not be true but is plausible and worth examining.