Social Work Research – Purposes, Classifications & Process (Comprehensive Notes)
Narrative Scenario: Jennifer & Paula – “Cart Before Horse”
Agency research committee charged with producing studies for mission fulfillment & accountability, aiming to demonstrate the effectiveness and responsible use of resources within their social work programs.
Paula, a committee member, is fixated on acquiring a specific stress-measuring gadget, prioritizing instrumentation before foundational planning. She struggles to articulate the overarching purpose of the study, highlighting a common pitfall in research design.
Jennifer, a more experienced researcher, prompts sequential thinking, emphasizing a logical progression essential for rigorous research:
Clarify purpose first: Researchers must begin by defining the study's objective. Is it to describe a phenomenon (descriptive), explain relationships (explanatory), assess effectiveness (evaluative), or explore unknown areas (exploratory)? This initial step dictates all subsequent methodological choices.
Then craft a precise research question: Once the purpose is clear, a focused, answerable research question can be formulated (e.g., “Are younger clients more likely to miss appointments, and if so, what factors contribute to this?”). This question guides the entire investigation.
Conduct literature review to frame concepts & methods: A thorough review of existing literature provides theoretical grounding, defines key concepts, informs variable selection, and helps identify established research methods or instruments.
Decide on study subjects (population & sample) and valid measurements: Based on the research question, identify the target population and the specific sample to be studied. Crucially, select measurement tools that are both reliable (consistent) and valid (accurately measure what they intend to).
Collect & analyze data → draw conclusions: Only after systematic planning are data collected using predefined methods, then analyzed using appropriate statistical or qualitative techniques, leading to evidence-based conclusions.
Moral: Never select instruments or specific methods before articulating the clear purpose and research question; “Don’t put the cart before the horse!” This maxim underscores the necessity of strategic planning over premature tactical decisions.
Big-Picture Goals of the Chapter
Heighten awareness of how research is characterized so social workers can effectively conceptualize, plan, and execute projects, knowing “where they are” in any stage of the research process. This understanding is crucial for both conducting and critically evaluating research.
Specific competencies:
Classify studies by purpose: Distinguish between descriptive, explanatory, evaluative, and exploratory studies, recognizing the unique aims and implications of each.
Distinguish program evaluation vs. intervention evaluation: Understand the differing scopes and targets of these two critical forms of evaluative research.
Differentiate quantitative and qualitative measurement: Grasp the fundamental differences in data types, collection methods, and analytical approaches associated with numerical vs. textual data.
Outline the entire research process from question → conclusions: Comprehend the logical flow of research, from the initial formulation of a research question to the final interpretation and dissemination of findings.
Four Purposes of Social Work Research
Descriptive Research
Objective: To portray, describe, or summarize characteristics of a population, situation, or phenomenon, typically focusing on one variable at a time to provide a clear snapshot without exploring relationships between variables.
Typical questions: What is the average age of MSW students in a program? What proportion of agency clients have preschool children? What is the prevalent income level of clientele served by a specific program? How many clients completed a particular service?
Analysis always concerns one variable at a time, even when multiple variables are collected. For example, if a survey collects age, gender, and income, descriptive analysis would report each of these individually (e.g., mean age, percentage male/female, median income).
Precision in measurement is paramount; constructs (e.g., “conservatism,” “client engagement,” “housing stability”) must be clearly and unambiguously defined (operationalized) before seeking tools or methods to measure them consistently and accurately.
Explanatory Research
Objective: To explain phenomena by establishing cause-and-effect relationships or examining associations among two or more variables. This type of research seeks to understand why something occurs.
Example question: “Do regular exercisers suffer fewer minor illnesses compared to non-exercisers?”
Requires specific variables: defined participation in regular exercise (independent/predictor variable) and frequency of minor illnesses (dependent/outcome variable).
Statistical tests are used to judge whether observed group differences or correlations are statistically significant, meaning they are unlikely to have occurred by chance, thus suggesting a relationship.
Key element: A conceptual framework/theory that ties existing literature to the chosen variables and postulates expected relationships. This framework provides the logical underpinning for hypotheses and helps interpret findings. For instance, a stress-buffering theory might suggest that stressors lead to stress, but social support can moderate (reduce the impact of) that stress.
Evaluative Research
Objective: To assess the success or failure of programs, interventions, policies, or practices on specified outcome metrics. It determines if something is working as intended.
Focus here: Primarily Intervention evaluation (assessing the effectiveness of specific services delivered to individuals or groups).
Examples:
Measuring pre- and post-therapy depression scores for clients receiving cognitive behavioral therapy to determine if the intervention reduces symptoms.
Comparing students’ grades during a tutoring semester versus their grades before tutoring to assess academic improvement.
Tracking housing acquisition rates within months for clients in a housing program and comparing it against a national average or a target benchmark.
Critical issue: Causation. It is essential to determine whether observed improvements or outcomes are directly due to the services or interventions provided, or if other external factors (e.g., natural recovery, concurrent life changes) are responsible. This often necessitates robust research designs.
Exploratory Research
Objective: To probe under-studied or new phenomena; to generate initial insights, hypotheses, or theoretical frameworks in areas where existing knowledge is scant or nascent. This research is often a first step in understanding a problem.
Example: Investigating the potential stages of adolescent moral decline in a specific community where minimal existing literature or theoretical frameworks are found. This could involve interviewing adolescents, parents, and community leaders.
Often paired with qualitative methods due to their inherent flexibility, ability to capture rich, in-depth narratives, and capacity to uncover unanticipated themes and perspectives.
Key issue: Do the chosen methods truly “capture the essence” or comprehensively represent the nuanced and potentially unknown dimensions of the theme under investigation? This requires careful design to ensure depth and breadth of exploration.
Scope Within Evaluative Research
Program Evaluation (Broad Scope)
Examines multiple facets of an entire service system, program, or agency. It is a comprehensive assessment covering various aspects from implementation to outcomes.
Child abuse prevention program example: A program evaluation would assess not just client outcomes but also the quality of intake assessments, adherence of staff to service standards, staff training effectiveness, and the overall drop in repeat abuse incidents across the program's lifespan.
Systems terminology preview for conceptualizing program components:
Input: Characteristics of clients or resources entering the program (e.g., client demographics, presenting problems, staff qualifications, funding).
Process: The specific services, activities, or interventions delivered within the program (e.g., number of therapy sessions, quality of group discussions, fidelity to a treatment model).
Output: The direct, countable products of the program's activities or services delivered (e.g., units of service provided, number of clients served, attendance at workshops, number of referrals made).
Outcome: The changes or effects on clients or the community resulting from the program (e.g., reduction in depression symptoms, improved parenting skills, increased housing stability, decreased recidivism).
Efficiency: The relationship between program costs and the benefits or outputs achieved; often expressed as cost per output or cost-effectiveness.
Depression service vignette illustrating program evaluation:
Client eligibility (Input): A program might analyze if clients served are truly eligible (e.g., county residents with clinical depression).
Service delivery (Process): Review actual number of CBT sessions received (e.g., one-hour sessions vs. planned ), session quality, and therapist adherence to protocol.
Staff credentialing (Process Input): Verify if therapists are appropriately credentialed (e.g., LCSW).
Client outcome (Outcome): Measure the actual reduction in depression symptoms (e.g., by 35 ext{%} based on a standardized scale).
Cost analysis (Efficiency): Calculate the cost per client or per session (e.g., 800 ext{ per client for 8 sessions}201118 discharged patients from one hospital against adherence rates from other hospitals to evaluate the effectiveness of a specific discharge protocol.
Often equated with evaluating one’s own practice as a clinician, where therapists systematically track and assess the outcomes of their direct work with clients to inform and improve their clinical effectiveness.
Measurement Approaches
How data are translated from observations or responses into quantifiable or descriptive information.
Quantitative Measurement
Variables are captured as categories (e.g., male/female, marital status: single, married, divorced) or numbers (age in years, Beck Depression Inventory scale score, number of missed appointments). These measurements are typically pre-defined and standardized.
The researcher predetermines the categories or numerical scales (e.g., a Likert scale with options like: strongly agree, agree, neutral, disagree, strongly disagree). This allows for systematic data collection and statistical analysis.
Well-suited for descriptive, explanatory, and evaluative studies; it facilitates hypothesis testing by allowing for statistical comparisons and the identification of relationships between variables.
Qualitative Measurement
Data are primarily words: rich descriptions from interview transcripts, detailed open-ended survey responses, narrative observer field notes, or content from documents. This approach seeks depth and understanding of meaning.
The participant, rather than the researcher, typically chooses the wording and expression, providing high flexibility and allowing for emergent themes and unforeseen insights to arise.
A natural fit for exploratory studies or inquiries into complex social processes, lived experiences, and subjective meaning-making, as it allows for in-depth understanding of individual perspectives and contexts.
USC Libraries definition (paraphrased): Emphasizes understanding qualities, meanings, and socially constructed realities. It focuses less on numerical quantities or establishing precise causal measurements and more on interpretation and context.
Matching Purpose & Measurement (Rule of Thumb in Text)
Traditionally, descriptive, explanatory, and evaluative research purposes are often illustrated with quantitative measurement examples due to their focus on generalizability, statistical relationships, and outcome measurement.
Conversely, exploratory research is frequently illustrated with qualitative examples due to its emphasis on discovering new insights, exploring complex phenomena, and understanding individual experiences.
However, it's crucial to note that any purpose can employ either approach, or even mixed methods. The alignment is primarily pragmatic and based on what method best answers the specific research question, not a mandatory rule. For instance, a descriptive study could use qualitative interviews to describe the experiences of a specific group.
The Research Process (4-Step Model)
Step 1 – Develop Research Question & Knowledge Base
Originates from a blend of practitioner curiosity (e.g., observing patterns in client no-shows, noticing gender differences in treatment outcomes, seeking to understand client satisfaction) and gaps identified in existing knowledge.
Tasks:
Phrase clear research question tied to purpose: The question must be specific, answerable, and directly reflect the chosen research purpose (descriptive, explanatory, evaluative, exploratory).
Conduct literature review: A comprehensive review is essential to:
Define variables precisely: Understand how key concepts have been defined and measured in prior studies, helping to craft clear operational definitions for the current study.
Surface theories predicting relationships: Identify existing theoretical frameworks that can explain anticipated relationships between variables, providing a foundation for hypotheses.
Identify appropriate measures: Discover validated instruments, scales, or qualitative approaches previously used to study similar phenomena, helping to ensure the scientific rigor of data collection.
Sample classification exercise from text:
Exercise ↔ minor illness? ⇒ Explanatory (seeks to understand the relationship or effect).
Valuable manager traits? ⇒ Descriptive (aims to characterize or summarize attributes of managers).
Stressors ↔ stress? ⇒ Explanatory (investigates causality or association).
Social support ↔ stress? ⇒ Explanatory (explores how one variable influences another).
New Horizons depression pre/post? ⇒ Evaluative (assesses the impact of an intervention or program on an outcome).
Step 2 – Determine Study Methods
This step involves making concrete decisions about how the study will be conducted.
Population: Define the entire universe of individuals, groups, or entities relevant to the research question (e.g., all humans, all bulimic clients in a specific region, all social work agencies in a state).
Sample: Select a subset of the population to study. The sampling method used (e.g., random sampling, convenience sampling) dictates the extent to which findings can be generalized back to the larger population (external validity).
Measurement: Choose specific, validated instruments or approaches to collect data on the variables defined in Step 1 (e.g., the Beck Depression Inventory for depression, a structured interview protocol for qualitative data). Ensure the chosen tools are reliable (consistent) and valid (accurate).
Design (especially for evaluation): Select the overall structure of the study to minimize bias and allow for valid conclusions. Examples include one-group pretest/posttest designs, quasi-experimental designs, or randomized controlled trials (randomized group comparisons), which are crucial for establishing causality.
Step 3 – Collect & Analyze Data
This is the execution phase where raw information is gathered and processed.
Data: Individual pieces of information collected about each study participant or unit (e.g., an individual's age, their specific depression score, their response to an open-ended question).
Collection protocol could be structured as:
A one-off questionnaire administered at a single point in time to gather cross-sectional data.
Repeated administrations of measures (e.g., pre-intervention and post-intervention) to the same clients to track changes over time (longitudinal data).
Statistics chosen per question type:
Descriptive statistics (e.g., frequencies, percentages, means, medians) are used to summarize and describe the characteristics of the sample or variables.
Inferential statistics (e.g., t-tests, ANOVA, regression, correlation coefficients) are used to make inferences about the larger population based on sample data and to judge statistical significance.
Statistical significance gauges whether observed findings (e.g., a difference between groups, a relationship between variables) are unlikely to have occurred by chance alone. A finding is statistically significant if there's a low probability ( p < 0.05 n = 1140 ext{%}34 ext{%} higher than those of the non-tutored group (Group B).
This design effectively demonstrates the power of random assignment (which equates groups at the outset, minimizing pre-existing differences), the use of comparison groups (to isolate the effect of the intervention), and a clear outcome assessment to provide strong evidence of effectiveness.
Additional Key Concepts & Terminology
Accountability: The ethical and administrative imperative for social service agencies to provide evidence that they are fulfilling their mission, utilizing resources responsibly, and making a measurable positive impact on their clients and communities. Research is a primary tool for demonstrating this.
Efficiency: A measure of how economically a program or intervention delivers its services, typically calculated as Cost per Output. = rac{ ext{Total Cost}}{ ext{Units of Service}} rac{ ext{Total program cost}}{ ext{Number of therapy sessions offered}} rac{ ext{Total cost of housing program}}{ ext{Number of clients housed}}800 ext{ and provides }8 ext{ sessions, the cost per hour is }.
Input → Process → Output → Outcome: A widely used systems model for understanding and evaluating program components. Inputs are the resources and client characteristics entering the program. Processes are the activities and services delivered. Outputs are the direct, quantifiable products of these activities. Outcomes are the changes or benefits experienced by clients or the community as a result of the program.
Convenience Sample: A sample consisting of participants who are readily and easily available (e.g., students in a class, clients currently in an agency). While practical, it severely limits external validity because the sample may not be representative of the broader population, making it difficult to generalize findings.
Random Assignment vs. Random Sampling:
Random Assignment: A process used in experimental designs where participants are randomly allocated to different groups (e.g., treatment vs. control). This technique primarily enhances internal validity, ensuring that any observed differences between groups are likely due to the intervention rather than pre-existing differences.
Random Sampling: A statistical sampling method where every member of the population has an equal chance of being selected for the study. This technique primarily enhances generalizability (external validity), allowing findings from the sample to be confidently applied to the larger population.
Theory / Conceptual Framework: A graphic or narrative model that systematically links a set of concepts and postulates expected relationships among variables. It guides the selection of variables to study, informs the hypotheses, and provides a lens through which to interpret findings. Theories offer predictive power and explain why certain relationships might exist.
“Cart Before Horse” Maxim: A crucial guiding principle in research stating that methodological decisions, particularly the selection of measurement instruments or specific data collection techniques, must follow the clear articulation of the study's purpose and specific research questions. Premature focus on methods without clear objectives leads to unfocused or irrelevant research.