Discussion: Choosing and Using Single-System Design
Many people receive their first introduction to the scientific method in their early school years. The first experiments which students undertake typically involve plants, chemicals, or small animals in a tightly controlled experimental environment. These experiments enable students to establish a relatively clear cause-and-effect relationship between the outcome of the experiment and the manipulation of the variables.
As soon as a researcher introduces a human element, proving a cause and effect relationship becomes more difficult—as the researcher cannot enact total control of another person even in an experimental environment. Social workers serve clients in highly complex real-world environments. Clients often implement recommended interventions outside of social workers’ direct observation. Yet, evidence-based research calls for social workers to establish cause-and-effect relations between selected interventions and client outcomes as much possible. To meet this challenge, social workers must understand the study designs available to them and all of the variations of that design that can increase the rigor of the experiment and improve the likelihood of verifying a cause-and-effect relationship.
In this week’s case study, you will decide whether the social worker in the case study has appropriately chosen a single-system (subject) design and implemented it in such a way that it can be considered an appropriate example of evidence-based research. (Walden University, Discussion 3, 2021)
WHAT YOU WILL READ THIS WEEK:
Plummer, S.-B., Makris, S., & Brocksen S. (Eds.). (2014b). Social Work Research: Single Subject. Social work case studies: Concentration year. Laureate International Universities Publishing, pp. 70–72. [Vital Source e-reader].*** This is the case study regarding George, which we will use for this week\\\’s topics.
Pay special attention to the following learning resources:
Dudley, J. R. (2020). Social work evaluation: Enhancing what we do (3rd ed.) Oxford University Press. In this text, read the following section in Chapter 9: “Is the Intervention Effective?” under Outcome Measures: “Client Satisfaction & Effectiveness” through “Target Problem Scale” (pp. 236-244) *** If you are using the 2nd edition of this text (Dudley, 2014), contact me for comparable pages in the 2nd edition. You will need this information for the discussion.
Corcoran, K. & Hozack, N. (2010). Locating assessment instruments. In B. Thyer (Ed.), The handbook of social work research methods (2nd ed., pp. 65-74). Sage. *** A PDF version is easily accessible in the learning resources.
Mattaini, M. A. (2010). Single-subject studies. In B. Thyer (Ed.), The handbook of social work research methods (2nd ed., pp. 241-273). Sage. *** A PDF version is easily accessible in the learning resources.
Skim the other resources in the learning resources for Week 3.
WHAT YOU WILL LEARN THIS WEEK:
This week, we will explore research designs that are common in social work research, focusing on single-subject research designs and reviewing the multiple-baseline-over-target-behaviors single-subject design in detail. Single subject research designs are also called single case designs and single-system designs. I will use these terms interchangeably.
First, I will describe features of single-subject research designs. These features apply to ALL single-subject research designs.
Second, I will review different types of single-subject research designs, providing the advantages and disadvantages of each, with a special focus on the multiple-baseline-over-target-behaviors research design. The assigned readings in Dudley (2020) and Mattaini (2020) cover similar material. The material that I present below is OPTIONAL, but you may find it useful in organizing your thoughts for the discussion.
WHAT YOU WILL DO THIS WEEK
You will evaluate a research plan conducted by the social worker, Chris, in the assigned case study, answering questions to determine the type of design that is used and the appropriateness of the research methods.
Everyone, take a deep breath – read the following information carefully if you are not familiar with research designs and methods. You will use your current knowledge and experience as well as the information in this post for the questions, which you must answer as a researcher, not as a provider of clinical or social services. We will assume that Chris (the social worker in the case study) will be providing those. Your job is to think like a researcher. I am available to walk you through the process.
COMMON FEATURES OF SINGLE SYSTEM RESEARCH DESIGNS
What is a single-case research design, and how is it different from designs where groups are compared to each other?
1. In single case designs, the focus is on a single unit of study, such as an individual or a small group, that is compared only to itself. Comparing the system or individual to themselves is the key factor that identifies a study as a single-case research design.
In other words, the individuals or the system receiving the intervention are their own controls. “Unlike true experiments where the researcher randomly assigns participants to a control and treatment group, in single subject research the participant serves as both the control and treatment group.” (Fraenkel and Wallen, 2006, p. 318).
2. The second essential element of the single case research study is the use of the visual graphics (usually line plots) to show change by recording the frequency or magnitude of the target problem before, during, and after the intervention.
What’s the purpose of a single-case design, and will I ever use one?
Social workers often use elements of single case designs in their work with clients, although they may not realize it. Our practice requires us to observe the effect of our interventions on clients and the achievement of their goals. If it does, we have identified an effective intervention for them, but if it is not effective, we try something else. By performing these observations systematically, we are conducting some aspects of single subject research.
In a clinical practice setting, the effectiveness of interventions can most readily be evaluated by SCDs [single-subject research, also termed single-case designs]. Additionally, given the fact that social workers are the largest providers of mental health services in the United States, it is not surprising that, more and more, SCDs are being used in the delivery of mental health services. (Georgoulakis, Zollmann, Pate, & Hallet, 2017).
Mattaini, in Thyer (2010), notes that
SSR [single systems research, also called SCD, single-case design] has particular strengths for social work research. SSR [that is, single systems research] focuses on the individual system, the individual person, the individual family, the individual neighborhood, [which is] typically the level of analysis of primary interest in social work. Furthermore, SSR allows detailed analysis of intervention outcomes for both responders and nonresponders, which is critical for practice because each client, not just the average client, must be of concern […and] rigorous SSR can be implemented in natural and near natural conditions, making it a practical strategy” (p. 243).
What are the requirements for anevidence-basedsingle-system design?
Citing Jomer, Carr, Halle, McGee, Odom, and Wolery (2005), Tankersley, Cook, and Cook (2008) assert that single case designs must include a thorough \\\”description of participants and settings, the dependent variable, the independent variable, baseline [measures of the frequency or magnitude of the problem with careful attention paid to] experimental control/internal validity, external validity, and social validity” (p. 525). These tasks must meet or exceed the same standards of scientific rigor in the same domains that are applied to experimental studies with groups to meet the gold standard of an evidence-based practice.
In addition to scientific rigor for research validity, Tankersley, Cook, and Cook (2008) propose additional standards to ensure that the conclusions of single-case research are evidence based:
(a) at least five studies that meet minimally acceptable methodological criteria, document experimental control, and have been published in peer-reviewed journals;
(b) the studies must be conducted by at least three different researchers across at least three different geographical locations; and
(c) the studies cumulatively include a total of at least 20 participants”(p. 525).
How do I evaluate a single-system research design?
These are the key questions that determine whether a study was conducted with a single-system research design:
1. Is the focus on a single individual, system, or group (such as a classroom) where the individual, system, or group functions as its own control without comparisons to other groups?
2. Is the target behavior defined clearly and assessed with quantitative measures at regular intervals?
Assessments usually include counts (frequencies) or estimates of the magnitude of the target behavior and displayed visually, often as line graphs. The primary sources of information for the assessment include these measures:
1. observations of the target behavior or condition
2. self-report from the client (most common when monitoring thoughts or feelings)
3. goal attainment scales (how much a desirable behavior or situation is increasing) and
4. target problem scales (measuring how much an undesirable situation is decreasing)
5. Clinical scales, goal-attainment scales, and target scales are commonly used to assess progress.
5a. Clinical scales are used to evaluate the presence and severity of psychiatric symptoms including depression, anxiety, ADHD, and dementia.
5b. Goal attainment scales are typically used to assess performance – how well or how often a behavior is performed, or if the person achieves a standard for success (yes/no)
5c. Target problem scales (also called target scales) are used when a goal is broken into intermediate steps (targets). The client’s level of success in achieving each target step toward the goal is monitored.
3 Are visual graphics such as line plots used to track change in a target behavior from baseline to the end of the study so progress can be easily visualized?
In addition to these, all research must answer the following questions:
1. Is the intervention appropriate for the problem and the developmental leval or capacity of the participant? Does the intervention conform to the core values of the Code of Ethics of the National Association of Social Workers (that is, service, social justice, dignity and worth of the person, importance of human relationships, and integrity)?
2. Are the criteria for success or effectiveness clearly understood and accepted?
3. Are the researchers and clinicians competent to conduct the research?
4. Have all appropriate ethical practices and oversight from an IRB been implemented?
What types of single-system research designs are used in social work?
At least four different single subject designs are common in social work. They usually fall into one of the following categories:
1. AB (baseline + intervention phases). One target behavior or condition is observed (A) and then ONE intervention (B) is applied to address it. The status of the target problem is compared before and after the intervention to learn if the intervention was helpful. In our work, this is probably the most common type of single-case study.
1a. Concurrent AB is a design in which several target behaviors are observed before an intervention and then interventions are applied to the all the problems at once, making it impossible to tell which interventions are (or are not) influencing the target behaviors. THIS IS IMPORTANT TO REMEMBER!
It\\\’s like throwing all your interventions at the problem, all at the same time. The difficulty with this design is that you cannot know which intervention is causing the change or which intervention is working better than another.
This design may be used when multiple problems are urgent and there is no time available to see how an intervention impacts one problem before applying an intervention to another problem. BUT, since the interventions are applied without time allowed between them to determine how each one responds to the treatment, concurrent AB design has poor external validity. In other words, we do not know if the interventions would work (or would fail to work) for anyone else.
Concurrent AB Design
Notice that the treatment (B) is being applied to all three behaviors at once. Can we be sure that the intervention is having the effects? Could it be that improvement in one area automatically improves another outcomes in another area? Think about improvement in sleep for adolescents. Improving sleep might also improve behaviors, attendance, and achievement at school – not because of an external treatment.
The concurrent AB design does not allow for generalizations (external validity) because we do not know which intervention is influencing the changes in the target behaviors. Avoid if at all possible!
2. ABC (baseline + intervention 1 for a set period; then intervention 1 is discontinued and replaced by intervention 2), sometimes called multi-component intervention design. The target behavior is observed during the baseline period (A) followed by application of the first intervention (B). Then the first intervention (B) ends and is replaced with a second intervention (C). These designs are very common in mental health where a client will provide baseline information during an assessment (A), an intervention, such as counseling, is provided (B) but counseling is discontinued due to insurance or other issues and replaced by a different approach (C), often medications.
The problem with the ABC design is that the first treatment might have a delayed effect. As a result, the improvements seen with Treatment C might actually be delayed improvements from Treatment B. Several studies would have to be conducted to determine the actual effect of each treatment ( B and C).
3. ABA or ABAB (baseline + intervention + withdrawal of intervention; with ABAB, the intervention is then restored). These designs can provide a stronger test of the effect of the intervention. If the problem returns when the intervention is discontinued but improves when the intervention is provided again, it suggests that the intervention is producing or contributing to the change. The ABAB design poses some ethical issues for social workers because it may not be possible to remove an intervention that appears to be helping a client to progress, but these designs sometimes occur naturally when clients discontinue an intervention on their own and then return to the intervention at a later time.
For example, a client struggling with a mental health issue who does well with medication may decide that he no longer needs the medication, only to observe that the symptoms of his mental illness are returning when the medication is discontinued. If he finds the symptoms remitting a second time when he resumes the medication, he has imposed an ABAB single case design on his own problem without realizing it. These results enable the social worker and the client to conclude that the medications make an important difference in his well-being.
Both ABA and ABAB are called withdrawal designs because the treatment (B) is provided after the baseline period (A) but then withdrawn. In the ABAB design, the treatment (B) is provided a second time.
These designs help researchers to determine if the changes and improvements in the participants\\\’ conditions are due to the treatment or to an external influence. In some cases, however, withdrawal of treatment is unethical, making this design inappropriate for any research involving vulnerable participants, serious health issues, and safety.
5. MB: multiple-baseline-over-behaviors. Pay close attention to these details because they pertain to the case study for this week.
In this design, baseline data are collected on two or more behaviors at the same time to control for carryover effects and learn how effective the intervention is with each target behavior separately. Only after baseline data on all behaviors are collected is the treatment administered – but the treatment is provided sequentially to the behaviors. This design is applied in steps with each behavior being addressed in order.
The process requires baseline information to be obtained on each target behavior or condition. Then the intervention is applied to one target problem while the baseline data continue to be collected on the other target problems but are not treated. After the first behavior has been treated with the intervention, the second behavior also is provided the intervention for a period while baseline data continue to be collected on the third behavior. When both the first and second targets have been treated, the intervention is added to the third behavior, so that the intervention is now applied to all three behaviors.
Cooper, Heron, & Heward note that \\\”interventions cannot be applied to the next behavior until the previous behavior change [if any] has been established\\\” (2007, p. 214). After changes to the first target problem are stable, the intervention can be applied to subsequent target problems in sequence, holding untreated problems at baseline while the intervention continues on the first target behavior and is added to the second behavior. When the responses of the second behavior to the intervention are stable, the intervention continues on the first and second target behaviors, and is applied to the third behavior which has been held at baseline while the other behaviors were treated. In this manner, the researcher can observe how each problem responds to the intervention.
Differences from other single case research designs.
How is the multiple-baseline-over-behaviors research design different from the ABC design? In the multiple-baseline-over-behaviors design, the intervention continues to be applied to the first target when the intervention is applied to the second target. In the ABC design, there is one target behavior and two interventions. The behavior is A, the first treatment (B) is applied, then stopped and replaced with intervention #2 (C). The problem with this method is that there could be carryover effects from Intervention #1 that occur when Intervention #2 is applied, making Intervention #2 appear more effective than it actually is, because the behavior is responding to the earlier intervention.
How is the multiple-baseline-over-behaviors design different from the concurrent AB design? The concurrent AB design does not allow time for the intervention to be applied in sequence to a set of target behaviors. The intervention is applied to all the behaviors at once after a short baseline period. As a result, it is not possible to estimate how the intervention affects each target behavior differently since there may be carryover effects from one target behavior to another.
For example, consider a high school student who cannot stay awake in their first period algebra class. Let’s assume that the social worker decides to offer a reward system for getting up on time, eating breakfast, and paying attention in their first period algebra class. The reward system is the intervention.
The social worker applies the reward system to all the behaviors at once. In two weeks, they find that there is an improvement in the student’s ability to pay attention in the algebra class. Is it because of the reward system or because the student now has eaten breakfast and is waking up on time? There is no way to know how much of an effect the intervention has on this behavior because the true cause of the change is not clear.
Multiple baseline designs can address different target problems in the same individual (most common use in social work) and are also used to address the same problems by other clients, which is called multiple baselines over subjects.
Multiple baseline across behaviors design – each plot refers to a different target behavior. Note that the baseline period is different for each target behavior.
In our example, the multiple baseline design helps to determine if an intervention is effective with several different target problems. The advantage of the multiple-baseline-over-behaviors design is that it controls for the effect of external events or \\\”spill-over\\\” effects of change from one target behavior on other target behaviors.
What you will DO
Imagine that you are the social worker assigned to a research project to assist your client, George.
Your task is to conduct a research project to learn which interventions are most effective for improving George\\\’s functioning in three areas: home safety, personal care, and mobility.
Your goal is to evaluate the research design and determine if the research and assessment methods will answer Chris\\\’s research question on the effectiveness of the intervention (case management with solution-focused and task-centered approaches) on the behaviors targeted in this study. You will also provide recommendations for improving the study if Chris and the research term intend to submit the study to the evidence-based research registry.
To prepare for this Discussion, review the case study of George (Social Work Research: Single Subject) and the criteria for using single-system research designs in this week’s learning resources and in the information provided above.
After reading the case in the Sessions book, you have some knowledge of George\\\’s situation. Those factors and some additional assumptions include these factors:
1. George is 87
2. George lives alone since the death of his wife two years ago, but has adjusted to his loss.
3. George has experience with science and is interested and able to providing accurate observations of his behaviors for the study.
4. George has mild cognitive decline.
5. Despite his cognitive decline, he is competent to give consent and is willing to participate in the study.
6. You are an experienced and competent social work researcher.
Results from your study:
After implementing the interventions, you have the following data (look closely at the changes in each target behavior following the intervention for that behavior)
Plot of George\\\’s Data, Showing Changes in Behaviors as the Intervention is Applied to a new Target Behavior at Two-Week Intervals (Weeks 3, 5, and 7)
YOUR POST: By Day 3, (Wednesday, 11:59 PM, Mountain Time,) provide your evaluation by answering the questions listed below. Please label your entry with \\\”Original Post\\\” to identify your evaluation.
Keep in mind that single case/system/subject designs are applied when the researcher wants to discover the impact of intervention(s) on one person who serves as their own control.
Then answer the following questions:
1.Does the approach used by the social worker with George meet the criteria for a single case design? Explain your decision. (1-2 sentences).
2. Explain the research goal by describing if the social worker is assessing the effect of one or more interventions on one target problem or the effect of one intervention with several target problems. (1-2 sentences).
3. Describe whether the intervention is introduced to all George’s problems at the same time (that is, the concurrent AB research design), or if it is applied to each target problem in sequence, staggering the introduction of the intervention to the target behaviors (multiple-baseline-over-behaviors research design). Identify the research design and the corresponding target behavior(s). (1-3 sentences).
4. “As part of the … process, the case manager used clinical rating scales that were adapted from the task-centered model [to address each of the target problems]” (Plummer, Makris, & Brockson, 2013, p. 71). Explain whether you support the use of clinical rating scales as an appropriate assessment method for the target behaviors in this study. Clinical rating scales are not defined in this vignette, but I assume they would assess both emotional and physical responses to the interventions and would rely on self-reports. (1-2 sentences).
5. Describe two additional assessment methods that would be useful for determining George’s status on the three target areas being addressed.
6. Using your own experience, what additional information would you would seek to assess his progress on the target behaviors? (Remember that you are wearing your research hat in this situation, not providing therapy or assessing other aspects of George’s physical, cognitive, or emotional condition.) What method would you use to obtain information on his progress? Provide your rationale. (2-3 sentences).
7. From the details presented in the study, explain whether you can determine if the intervention and study have been conducted ethically. (1-2 sentences).
8. Notice the trends in the CHANGES in George’s scores in the tables in the case study and interpret the data by indicating whether the trend for each behavior shows improvements (significant, moderate, or small), no improvements, or worsening of the target problem. Respond by selecting responses in these statements:
In evaluating the social worker’s use of an intervention consisting of case management and solution focused and task-centered methods, I observed (select one: significant improvement, moderate improvement, slight improvement, no improvement, worsening problems) in George\\\’s home safety behavior.
I also observed (select one: significant improvements, moderate improvements, slight improvements, no improvements, worsening problems) in his personal care behavior and (select one: significant improvements, moderate improvements, slight improvements, no improvements, worsening problems) in his mobility.
9. Based on your interpretation, write a conclusion, stating whether this intervention appears to be effective, promising, or ineffective and provide a rationale, using the format below and filling in the blanks, as appropriate:
These results suggest that the selected intervention was (select one: effective, promising (evidence of effectiveness is present but small), or ineffective) for ensuring home safety, (select one: effective, promising (evidence of effectiveness is present but small), or ineffective) for improving personal care, and (select one: effective, promising (evidence of effectiveness is present but small), or ineffective) for increasing mobility. My decision is based upon _____ [draw your conclusions from the graph for George\\\’s response to the intervention]. (2-3 sentences).
10. Make a recommendation based on your analysis, following this format:
Given the importance of addressing these issues in my community, scarcity of funding and accompanying need for careful allotments of the social service budget at this time, I (would/would not) support this program of using the task-centered solution focused case management intervention for ____ (home safety, personal care, mobility – choose all, some, one, or none).
11. Provide one recommendation for an improvement to this study or the description of their case study, \\\”should Chris and her colleagues wish to submit the study to the evidence-based practice registry. Include a rationale for your recommendations” (Walden University, 2021). Use the criteria listed above by Tankersley, Cole, and Cole (2008) for establishing an intervention as evidence-based for your recommendations.