To test the effectiveness of using a dab of honey to prevent infection in a cut, parents of children in a childcare facility were instructed to wash the wounds of their children with soap and water and then place a dab of honey on the cut. each cut was then monitored for infection. at the end of the study, the percentage of the twenty eight children who had cuts that became infected was computed. does this study have randomization?

Question

Social Studies

Lane Atkins

12 January, 11:26

To test the effectiveness of using a dab of honey to prevent infection in a cut, parents of children in a childcare facility were instructed to wash the wounds of their children with soap and water and then place a dab of honey on the cut. each cut was then monitored for infection. at the end of the study, the percentage of the twenty eight children who had cuts that became infected was computed. does this study have randomization?

+5

Answers (2)

Know the Answer?

Chaim Mendoza · Answer 1

This is this not a valid experiment because there is no randomization. T here is no blinding as parents may use more care and those monitoring for infection may be less stringent in their diagnosis. There is no control group that gets a placebo for comparison. Parents may be more careful of the cuts of their children since they know they are in an experiment. And t he results may not apply to all children since only those children in a day care facility participated.

Kayla · Answer 2

If an experiment is conducted in order to assess the effect of a changing variable both a treatment and control group are needed, the first group receives the treatment and the later does not. Individuals have to be randomly sent to one of the two groups

In this example the treatment variable is a dicotomic one: wash the wound with honey (=1) or not (=0). We want to asses if it affects negatively the probability of getting a would infection. It is necessary to rule out other factors that can affect this probability, other than the treatment variable, such as the healthcare access, the hygene conditions, the family income, the number of siblings, etc. This is why we need two groups, because the control group is affected by all the variables except the treatment, and the treatment group are affected by all variables (including the treatment). Hence, the final conclusion on the effect of the treatment is reached by comparing the results in the two groups.

If there is only one group, there is not a random assingment of the treatment whose effect we want to quantify and therefore the results will be contaminated by a selection bias.