All your questions are answered
Aphasia Access is committed to expanding the base of quality aphasia research available on the Life Participation Approach to Aphasia. These aphasia research resources are designed to assist those new to conducting research studies that may be appropriate for peer review at some point.
single case studies & experimental designs
#1: Getting Started
Aphasia Access’ Education and Research Committee, is committed to expanding the base of quality research available on the Life Participation Approach to Aphasia. To that end, the committee has begun releasing a series of “Frequently Asked Questions” designed to assist those new to conducting research studies that may be appropriate for peer review at some point.
Below is the first in the FAQ series – Getting Started. Questions may be directed to firstname.lastname@example.org.
Please Note: Aphasia Access does not offer this material in lieu of expert advice.You are encouraged to familiarize yourself with the legal, ethical, and scientific practices that apply in your local setting or practice. If you have a research project in mind but have limited experience with research ethics or design, we recommend you connect with someone who has the relevant expertise.
I hear a lot of talk about the complexities of getting “IRB approval” for research. What exactly is an IRB?
All human subjects research is subject to local, state, and federal regulations governing the safe and ethical treatment of participants. All universities, hospitals, and other entities that conduct grant-supported research have boards or committees that review proposed research projects to insure compliance with the relevant laws and statutes. Institutional Review Boards (IRBs) or Research Ethics Boards (REPs) are the most common such committees.
As a free-standing aphasia center, our activities do not come under the auspices of an IRB or other research ethics committee. What are my options?
If your research involves collaboration with a hospital- or university-affiliated researcher, your collaborator’s research ethics committee should be able to provide the necessary oversight. If you are not currently engaged in such a collaboration, it would be advisable to seek a local institutional or commercial board that could provide the necessary oversight. Another option is to form one’s own research eithics committee before planning a research project.
As a clinician in an aphasia center not currently affiliated with an IRB, when and how do I get started learning about research ethics and review?
If you are thinking of gathering outcomes or other data from the participants in your program that might one day be published, or if you plan to apply for grant funds to gather such data, the time is right to familiarize yourself with some of the many research-ethics training resources available online. For example, the U.S. National Institutes of Health offers a free training course at https://phrp.nihtraining.com/index.php. Also, many universities have training videos posted on YouTube.
My aphasia center routinely gathers data about participant satisfaction. After each program module, we obtain information about what the participants liked about the module, what they didn’t like, and what changes they would like to see. Does this qualify as research?
It depends. Research is systematic data collection aimed at contributing to generalizable knowledge by answering a question or testing a hypothesis of broad interest to the field. Research is different from data gathering for purposes of Quality Assessment and Quality Improvement (QA/QI). QA/QI, also known as program evaluation, is aimed at producing immediate improvement in a local program or model. If your center’s purpose in measuring user satisfaction is solely to improve its program offerings, the activity counts as QA/QI, not research. But if the activity is aimed at a broader scientific purpose, such as building the evidence base for an aphasia treatment or outcome measure, then it probably does qualify as research.
Do QA/QI activities require approval from an IRB or other research ethics board?
No. An aphasia center’s QA/QI activities are generally overseen by its Board of Directors or institutional administrators. They, along with the center’s consumers and clinicians, are stakeholders in the program’s success and the ones responsible for QA/QI oversight. IRB oversight is usually reserved for human subjects research. The IRB is not concerned with program improvements but rather with the ethical treatment of the human subject participants, including their voluntary participation and rights of confidentiality.
Does data-gathering for QA/QI require written informed consent from participants.
No. However, in the aphasia center context, it is considered good practice to involve participants in QA/QI decision-making and to inform them in advance of the extent and purpose of the QA/QI data-gathering.
How can I be sure that my activities are QA/QI and thus exempt from review by an IRB or other research ethics board?
This is known to be a challenging issue. There are no hard and fast rules, and it is not uncommon for QA/QI to generate research questions and evolve into a research study. The chart below offers some guidelines; but it is best to consult with the Chair of your research ethics board, who can help you think through the various factors and decide if an application is warranted.
If I’ve been tracking enrollment statistics and related data for purposes of program evaluation, can I later use those data to plan a research study, e.g., to estimate the size and characteristics of the available research sample?
Yes. As long as it doesn’t infringe on the privacy/confidentiality rights of those who contributed the data, most research ethics boards would consider this an acceptable use of the data. Moreover, most biostatisticians would consider it wise practice.
If I gather outcomes data for the purpose of program evaluation, can I later use those data to answer a research question, e.g., can I contribute them as part of a multi-site research study?
If the data were gathered for program evaluation, the program participants did not give informed consent to have their data used or shared in this way. So it’s best to consult with your research ethics board before proceeding. They may ask if the data in question can be linked in any way to the individuals who contributed those data. If the data can be linked back to the individuals, the ethics board will probably require that those individuals agree in writing to this proposed use of their data. Generally speaking, its best to consider at the start of the project whether the data might be used for other purposes, and plan accordingly.
#2: Single Case Studies and Single Case Experimental Designs
Aphasia Access’ Education and Research Committee, is committed to expanding the base of quality research available on the Life Participation Approach to Aphasia. To that end, the committee has begun to release a series of “Frequently Asked Questions” designed to assist those new to conducting research studies that may be appropriate for peer review at some point.
Below is the second in the FAQ series – Single Case Studies and Single Case Experimental Designs. Questions may be directed to email@example.com.
Please Note: Aphasia Access does not offer this material in lieu of expert advice. You are encouraged to familiarize yourself with the legal, ethical, and scientific practices that apply in your local setting or practice. If you have a research project in mind but have limited experience with research ethics or design, we recommend you connect with someone who has the relevant expertise.
Also note: This FAQ assumes familiarity with the issues discussed in Getting Started, concerning what constitutes research and how research differs from clinical data gathering. We recommend that you review that FAQ before beginning this one.
What types of questions can I address with single-case studies?
The types of questions are many and diverse. For example, such studies are useful for describing a novel clinical presentation or treatment response; recording an individual’s experience of living with aphasia; characterizing an individual’s deficits in relation to a cognitive model, or testing a specific hypothesis about the nature of the impairment or cause of recovery. Many of these purposes can be accomplished by careful observation and description alone; clinical or descriptive studies are of this type. However, where the purpose is to prove a specific causal or mechanistic account, something more is required. This something is experimental control, achieved by actively manipulating the participant’s environment or responses so as to isolate the factor of interest from those that might confound its description or measurement. Single-case studies that feature such experimental control are called single-case experiments or, alternatively, experimental case studies.
When are single-case studies considered “research”?
Case studies are considered research when their purpose is to contribute to general knowledge (i.e., knowledge that applies beyond the individual being studied). Although case studies, by definition, have limited external validity (i.e., results aren’t generalizable to other individuals), they can and do contribute to the development of general knowledge by allowing researchers to study phenomena and interventions in depth. For example:
- Non-experimental (i.e., clinical or descriptive) case studies contribute to general knowledge when they address an understudied area, help develop hypotheses and methods, or otherwise move the field in a new direction. An example is Holland’s (2006) exploration of some clinical, psychological and social-support factors that play a role in successful living with aphasia. The study reports in-depth interview data from each of 3 people with aphasia (i.e., 3 descriptive case studies), drawing out qualitative themes in each and using the similarities and differences among the three to frame hypotheses for future research.
- In another non-experimental case study, Haley et al. (2016) sought to advance general knowledge about long-term recovery from AOS by tracking speech-related changes in a single individual over a two-year period. The person with AOS was also a co-investigator and co-author, contributing video-recorded monologues that she produced over the period of the study to chronicle her own experience of recovery. The investigators analyzed these monologues using a combination of quantitative and qualitative measures. The quantitative measures were analyzed as a time series and tracked things like speech rate and speech-error frequencies, while the qualitative analysis classified and described the speech-related content and feelings expressed. The result is a rich, multi-faceted portrait of the dynamics of recovery from AOS, along with hypotheses and recommendations that move the field forward.
- Well-designed single-case experiments help build the evidence base in support of new aphasia treatments and approaches (Smith, 2012). For example, in a recent summary of research supporting the Script Training approach to treating aphasia, 13 studies were listed, of which 8 employed single-case experimental designs (Kaye & Cherney, 2016).
- Single-case experiments in the cognitive-neuropsychology style employ individually-tailored assessments to characterize the precise nature of an individual’s impairment. Research studies of this type aim to uncover patterns of individual strengths and weaknesses that inform theoretical models and/or motivate a particular approach to treatment. An excellent discussion and example can be found in Hillis (1998).
When are single-case studies considered suitable for publication?
The main requirements are that the study contributes to general knowledge, addresses an interesting question, and is well-designed, meaning that the methods are appropriate to the study question, the state of current evidence, and, where appropriate, the experimental hypothesis. Be aware that journals differ in their willingness to accept case study research in general, and non-experimental case-studies, in particular; so, it’s a good idea to investigate where similar studies were published before planning a submission.
Do case studies conducted for research purposes require prior approval by an IRB or other research review board?
Generally, yes; but review boards differ in how they treat such research. For example, some boards are willing to review a single, umbrella proposal covering a series of planned case studies, whereas others insist on a separate review for each. Some will exempt from review the retrospective publication of case-study data collected for non-research purposes, whereas others will not. To avoid unnecessary complications down the line, it’s best to consult with your local IRB board early in the process of planning case-study research.
I’ve noticed that in many publications, several case studies are reported together. Why is this?
Sometimes this is done to highlight contrasting patterns of performance that are of potential interest, theoretically or clinically. Sometimes it is done for reasons of replication, to show that the case-study finding of interest is not unique to a single individual. Either way, the impact of the study is heightened by reporting the multiple cases together. Replication serves a special function in the “multiple baseline across subjects” experimental case-study design (see #7). The idea here is that showing a predicted change from baseline to treatment in each of a succession of participants lends confidence to the internal validity (i.e., experimental control) for the experiment as a whole.
What are the essential features of a single-case experiment on treatment efficacy?
Experimental single-case studies control participant’s actions or environment in the service of uncovering causal relationships or mechanisms of recovery. In its simplest form, the purpose of single case experimental treatment designs is to demonstrate a causal relationship, at the level of the individual, between administration of treatment and change in the target behavior. In a typical such experiment, a time series design is used, wherein the target behavior is sampled repeatedly throughout distinct experimental phases that are controlled by the researcher. These phases may include daily probes that precede the start of treatment (baseline phase), are conducted during the treatment period (treatment phase), and continue after treatment has ended (maintenance phase). Results are summarized in graphs that plot the individual’s scores on the target measure across sessions and phases of the study. This enables visualization of how the outcome variable covaries with the administration of treatment and its removal. As most aphasia interventions are expected to have effects that outlive the treatment, one expects that relative to baseline, scores will improve during the treatment phase and remain high during maintenance. This classic time-series design accommodates many variations, some aimed at increasing experimental control (Thompson, 2006; and see Questions 7 and 8, for examples), others at growing the evidence base for promising clinical treatments. Towards the latter aim, researchers have proposed design modifications that accommodate stronger statistical analysis of single-subject treatment efficacy data (e.g., Howard, Best & Nickels, 2015) and that facilitate comparison and averaging of effect sizes across related studies (Beeson & Robey, 2006).
What is a multiple-baseline design?
Ans. It is a form of experimental single-case design that is common in aphasia research. One must always be mindful of the possibility that what appears to be a treatment-related improvement from baseline is simply a consequence of repeated testing on the outcome measure. Multiple baseline designs control for this possibility in different ways:
- Multiple-baseline across participants – Here, several similar participants are enrolled in the same experiment, at approximately the same time, and the length of the baseline phase is staggered across the participants/subjects. If in each case, the rise in outcome scores coincides with the delivery of treatment (rather than length of baseline) then we can be more confident that it is truly the treatment that’s causing the outcome variable to change, and not its repeated probing or some external event in the participants’ life. Note that although this design features several participants/subjects, it is still a single-case experiment because the data are not aggregated across subjects and the individual participants are essential components in the experimental design.
- Multiple-baseline across behaviors – This type of design is useful with treatments that are known to produce item-specific or behavior-specific effects. In such cases, the distinct items or behaviors are probed during all phases of the experiment, and treatment is applied sequentially to each, in staggered fashion. If improvement on behavior b coincides with the treatment of b, then we can be confident that b’s improvement is truly due to treatment, and not repeated probing or external factors. In this way, participants serve as their own controls.
Can you give an example from aphasia of a multiple-baseline, across-behaviors experiment?
The following example is noteworthy in two respects: First, because its multiple-baseline design controlled the scheduling of treatment across both behaviors and participants; and second, because the experimental intervention coupled an impairment-level treatment with a group treatment to achieve a more functional outcome.
- The study, by Clausen and Beeson (2003), involved 4 participants/subjects, each of whom had chronic, severe aphasia and very limited verbal or written production. All had previously demonstrated some capacity to re-learn the spellings of single words, one by one, by means of a copy-and-recall treatment (CART). The 4 were enrolled as participants in the CART + group-treatment experiment, which ran as follows: In separate sessions with each participant, CART was used to train written spelling of 5 successively trained “behaviors”, here, sets of words selected by the individual to be personally meaningful. In the group treatment sessions, the 4 participants together engaged in clinician-facilitated conversations that encouraged the use of their written words to ask and answer questions. Performance on the target word-sets was probed in each individual session and each group session. Visual and statistical analysis of the data showed that each participant demonstrated improved spelling of target words (individual session probes) and improved use of target words in conversation (group treatment probes). The authors concluded that for individuals with limited spoken language, CART plus group treatment can be a viable option for improving and encouraging the functional use of written words in everyday interactions.
Beeson, P.M., & Robey, R.R. (2006). Evaluating single-subject treatment research: Lessons learned from the aphasia literature. Neuropsychology Review, 16, 161–169.
Clausen, N. S., & Beeson, P. M. (2003). Conversational use of writing in severe aphasia: A group treatment approach. Aphasiology, 17, 625-644.
Haley, K.L., Shafer, J.N., Harmon, T.G. & Jacks, A. (2016). Recovering with acquired apraxia of speech: The first 2 years. American Journal of Speech-Language Pathology, 25, S687-S69.
Hillis, A.E. (1998). Treatment of naming disorders: New issues regarding old therapies. Journal of the International Neuropsychological Society, 4, 648–660.
Holland, A.L. (2006). Living successfully with aphasia: Three variations on the theme. Topics in Stroke Rehabilitation, 13, 44-51.
Howard, D., Best, W., & Nickels, L. (2015). Optimising the design of intervention studies: Critiques and ways forward. Aphasiology, 29, 526-562.
Kaye, R.C. & Cherney, L.R. (2016). Scripts templates: A practical approach to script training in aphasia. Topics in Language Disorders, 36, 136–153
Smith, J.D. (2012). Single-case experimental designs: A systematic review of published research and current standards. Psychological Methods, 17, 510–550.
Thompson, C.K. (2006). Single subject controlled experiments in aphasia: The science and the state of the science. Journal of Communication Disorders, 39, 266–291.