Single Subject research design

alternating treatments design types

1. alternating treatments design with no baseline 2. alternating treatments design with a baseline 3. alternating treatments design with a baseline and a final treatment phase

ABAB research design

A1: first baseline (usual care, no study intevention) A2: second baseline = A1 B1 first study intervention or treatment phase B2: second study intervention = B1

Phase A represents? Phase B respresents?

A: control (baseline) B: intervention

types of single subject research designs

ABAB design (withdrawal/reversal) alternating treatments design multiple baseline design

is AB or ABAB better

ABAB is better

three major types of single subject research design

ABAB, alternating treatments, and multiple baseline

In a single subject research design study, the researcher reported that the findings were clinically significant. What do the findings mean in terms of statistical significance? a.The findings were not statistically significant b. since the findings were clinically significant, they must be statistically significant too c.Cannot be determined

C. cannot be determined No relationships between clinical and statistical significance so that is why cannot be determined is correct (don't have information and depends on situation)

example of when it is not possible to use same dependent variable to collect data

IQ tests can't use that multiple times because it wont change on a daily basis IQ stays the same

momentary time sampling

Record (multiple) behaviors only at the end (or beginning) of each time interval. E.g., observe the client for 5s at (beginning or end of) 10-min intervals for every 1-hr. want to collect data for more than one person in a period of time

interval recording

Suppose you are going to collect data on the frequency of peer interaction of an 8-year-old autistic child in a special school. You collect frequency of his peer interaction for 5 days with a 30-min session per day. The results are as follows: •Day/Session 1, you observe the child interacts with his peer 21 times during the 30-min period. •Day/Session 2, 22 times •Day/Session 3, 25 times •Day/Session 4, 22 times •Day/Session 5, 23 times

Advantage ABAB design

able to demonstrate an experimental control because it requires the repeated application and withdrawal of a treatment decreases the probability that observed changes are the result of coincidental change in confounding variables

ways to reduce reactivity

allow clients to adapt to observations before formal data collection begins use unobtrusive observational procedures

use several measures to help reduce the effects of erros when the outcomes

are more difficult to measure (attention to task) display in many different behavioral forms (outcome for using self-feeding system)

violated assumption

cant run traditional statistics when person is the same for control and intervention phase math is not done right ex: 3-4/ 2 you could get 1 or -0.5 depending on the order you do math due to you assuming its done one way

difference between case report and SSRDS

case report does not collect data SSRDS collect data, manipulates independent variable

reactivity

clients may respond atypically (social desirability) as a result of being observed

example of SSRDS

collecting data on pain level daily on baseline, treatment, after treatment applied with a patient with a chronic disease

purpose of having a baseline

define the extent to which a subject requires treatment judge the effects of treatment against baseline establish a reliable estimate of confounding variables (control for normal variation)

rising baseline is not problematic if slope in treatment phases is

different

what is the problem with having more than once subject

difficult bc it is very intensive data collection on daily basis and takes months to do, so easier to use just one person

if the baseline is not improving then you do or dont need intervention?

do

if you are getting better during baseline then you do or dont need intervention?

dont

example of clinical significance

dont care about P value real change from one stage to next stage clearly ex: person at beginning only able to sit on side of bed, now he can walk to bathroom

conditions under which single subject (AB) research design are not appropriate

event or occurrence during the study period might have produced effects on the dependent variable and natural recovery effect are the major threats to the internal validity repeated measurement of the same dependent variables is not possible or invalid (sometimes not possible to use same dependent varibale to collect data)

observer bias

expectation (deliberate or unintended effects) for improvement might affect the objectivity of the observation

clinical significance

extent to which therapy moves someone outside the range of the dysfunctional to within the range of the functional category effect is real and genuine, palpable, noticeable in daily life ex: taking fewer meds

major characteristics of SSRD

focus intensively on behavioral response of one or a few individuals (observing behavior) involved systematic manipulation of an intervention to determine its functional/ cause and effect relationship with its outcomes the subject serves as his or her own control when comparing performances repeated measurement of the dependent variables over time under standardized conditions emphasis on clinical significance

for AB to be good what do you need A and B data to look like

going in opposite directions ex: is A is increasing and then B decreases you are good (A being pain is increasing, B being intervention made pain decrease) if A is decreasing and B is increasing then you are good (A being ability to recall words decreases, B the intervention causes recalling of words increases) good AB study

multiple baseline design types

multiple baseline design across subject multiple baseline design across settings multiple baseline design across behaviors

how many data point should you have in baseline and should they be sloped or stable

need a stable baseline of at least 3 data points to go on to intervention

in image is the baseline stable (the first part)

no it is not, need more data points to be stable not good enough to go onto intervention

observer drift

occurs when shifting standards overtime result from observer experiences or boredom> resulting in decrease consistency

data collect needs to happen

on a regular basis over time

ways to reduce observer drift

over train observers periodically train, recalibrate (periodically rate criterion videotapes)

operational definition of independent variable

provide detailed decriptions of the intervention exactly what it is

since there is only one person and they are acting as their own control group this make it what type of experimental study

quasi-experimental

example of operational definition of dependent variable using peer interaction

refers to a social relationship between age-peers such that they mutually influence each other

operational definition of dependent variable

refers to phenomena that are concrete, observable and measurable at some level delineates the boundaries of the response so that an observer can discriminate it from other responses includes some borderline or difficult examples of the responses and exclusion criteria

limitations of ABAB design

removal of a treatment may cause resentful demoralization when the performance gains cannot be easily reinstated after a withdrawal of treatment prone to carry-over impossible to reverse the change attrition of the subject limited generalizability only give clues to what is more than likely to happen in similar circumstances with similar subjects naturally occurring cyclic changes will be mistakenly attributed to the intervention behaviors have to occur frequent enough to be easily observable require relatively longer term commitment for data collection rules for interpreting visual analysis (subjective)

statistical significance

results obtained in analysis of sample data are unlikely to be caused by chance at some specified level of probability (alpha = level of significance= 0.05 or 5%

which is stronger? case report or single subject

single subject is stronger because it has more data points case report only has one data point so not strong enough

what does systematic manipulation mean (apply, withdraw, vary)

sometimes introduce intervention and then take it away and then introduce it again used for research design not clinical

the image is an example of

stable baseline has at least 3 stable points at end

types of dependent variables

staff assistance time number of time for physical assistance length of the meal percentage of food consumption weight change

are pain levels objective or subjective

subjective because you have to tell experimenter how you feel and that is different for everyone

standardize conditions for recording

times of day and setting across phases

examples of independent variables

types of feeding method self-feeding system or being fed

ways to reduce observer bias

use trained observers with stringent training criteria use objective measuring instruments use precise, low inference operational definition of dependent variable blind observers to the purpose of the study and experimental conditions

single subject research design study

uses experimental controls and repeated measures and attempt to find cause-and-effect relationships between variables make a systematic collection of objective reliable data take several measurements over time, before and after the intervention

conditions that increase confidence of interpreting a treatment effect

when baseline is stable or in a direction opposite of that predicted for treatment fewer number of overlapping data points between baseline and treatment phase more immediate effect following the introduction of treatment larger size of treatment effect greater number of times that an effect is shown both within (ABAB) and across subjects (multiple baseline across subjects)

instrumentation bias

when the measurement depends on observers who become more experienced over time if there is a change in the observer from baseline to the intervention phase

example of operational definition of independent variable

whether using microwave oven will reduce meal preparation time need to specify what model/type and power of microwave you use bc this affects cooking time

learning effect

with learning you can not go back to baseline after first AB because you have learned so your baseline will be too high up since you have improved

Is the image a strong ABAB design or not

yes it is very strong