"Researchers in Japan tested the effect of cannabidiol (CBD) i...

Question

"Researchers in Japan tested the effect of cannabidiol (CBD) in treating late teenagers with social anxiety disorder (SAD). Thirty-seven 18- to 19-year-old teenagers with SAD took part in the study. The patients were assigned at random to receive a daily dose of either CBD or a placebo for 4 weeks. Symptoms were measured using The Fear of Negative Evaluation Questionnaire and the Liebowitz Social Anxiety Scale at the beginning and end of the treatment. (Source: Frontiers in Psychology)
How could blocking be used in designing this experiment?"

How could blocking be used in designing this experiment?"

Accepted Answer

Hello and welcome back, everybody. Our next problem says a team of researchers is investigating the impact of a new concentration enhancing supplement on college students. A total of 40 students aged 18 to 20 agreed to participate. The students are randomly assigned to receive either the supplement or a placebo daily for 2 weeks. Their performance is assessed using standardized focus and attention tests at both the beginning and the end of the study. How could blocking be used in the design of this experiment to improve its accuracy? So, let's recall what blocking is, which is when we divide our overall population into blocks based on shared characteristics. These can also be called clusters. And then you randomly select a number of participants from each block, and note that the number of participants selected needs to be proportional to the blocks or clusters share in the population. So, if you, for instance, divide by age groups, you'd want to make sure that the number of participants you took from each age group is representative of what percent of the population is made up of that particular block. So, in this case, why would that improve the accuracy? Well, in general you do this to help eliminate confounding variables. So, other things that might affect the outcome. So we want to make sure that we're seeing the actual effect of the supplement and not some other factor that might affect their performance. So we could think about in this case, well, we've got an age range of only 2 years, so age blocks probably don't make too much sense. But what do we think of in terms of confounding variables? Well, we're thinking about a concentration enhancing supplement. Well, if you think of caffeine, it can help increase concentration. So perhaps we'd, so I'll put, for example, e.g. Divide By amount of caffeine consumed. So make sure that we don't have one group. Being overly represent or caffeine like hot lots of coffee drinkers overrepresented in a group. That might be selected in another way. So in this case, we might divide by, you know, non-coffee drinkers, moderate coffee drinkers, and heavy coffee drinkers, and then select a random sample from each of those groups to help eliminate the confounding variable of caffeine. So again, talking about, if we were doing a study of the impact of a concentration enhancing supplement on college students, we have a group of 40 students randomly assigned to receive the supplement or placebo. Assessing their performance with a focus and attention tests at the beginning and the end. How is blocking Going to be useful to improve the accuracy of this experiment. And the answer is that it helps eliminate. Confounding variables, because this divides the population into blocks or clusters based on shared characteristics. So something like caffeine consumption that might otherwise affect the results, we're sort of. Eliminating that confounding variable by making sure we are sampling people that consume different amounts of caffeine in a proportional amount. See you in the next video.

Key Concepts

Randomization

Blocking

Measurement Scales

Watch next