in this video we will be looking at experimental designs the types of experimental designs include the completely randomized design the randomized block design and the Matched pairs design in a completely randomized design each experimental unit is randomly assigned to a random group to receive a different treatment each unit in the same group will receive the same treatment and at the end of the experiment you will compare the results from each treatment show how a completely randomized design would be used in the following example a researcher wants to conduct an experiment to determine which environment is best suited for studying a library in one own's room or outside a total of 30 University students volunteer to participate in the experiment so we have three treatments the library in one own's room and outside since we have three treatments we will have three groups receiving the the treatments since there are 30 University students these 30 subjects will be randomly assigned into three groups to receive the treatments since there are 30 people each group will have 10 people and at the end of the experiment we would compare their results from each treatment let's say that the researcher believes that gender has an effect on the results in a case like this we would use a randomized block design a randomized block design doesn't immediately randomly assign experimental units to re receive treatments we first assign people into a block based on a characteristic that is expected to influence the response to the treatments in this example gender would be the blocking variable so first we separate our experimental units based on gender one block will be for females and one block will be for males there are 18 females and 12 males then the experimental units in each block will be randomly assigned to three different groups to receive a treatment so let's make a diagram out of this we are looking at a randomized block design and gender was the blocking variable we started off with 30 University students and we separated these students into two blocks one was for males and one was for females we determined that there were 18 females and 12 males we will have the females be randomly assigned to three different groups and each group will have six females and each group will receive different treatments for the males we we are doing the same thing except we have 12 males so we will randomly assign four males into each group and at the end of the experiment we will compare the results within each block the last type of experiment we'll talk about is the Matched pairs design in a matched pairs design we are only comparing two treatments by using the same or similar experimental units if we are using the same experimental units we could have an example like this what type of gasoline is more efficient type a gasoline or type B gasoline three cars will be used in this experiment describe how a matched pairs design would be used if we are using the same experimental units then each experimental unit will receive both treatments so we have three cars and each car will receive the two treatments one with type a gasoline and one with type B gasoline we would use the same car for each treatment and which treatment goes first will be random so for example we could drive with type B gasoline first then once all the gasoline is finished we would fill up the car with type a gasoline using the same quantity and at the end of the experiment we would compare the efficiency of the gasoline there are cases when we can't use the same experimental units for a matched pairs design and we would have to use similar experimental units for example a researcher is trying to determine whether or not sleep deprivation has an effect on test scores describe how a matched pairs design would be used well we could try giving both treatments to the same experimental units like we did in the other example in other words we could have the same person write the test on two different days one being sleep deprived and the other with sufficient sleep the problem with this is that we can't have the same subject write the same test because they would know what's on the exam when they write it the second time so that would be kind of unfair instead we could start off by pairing up students with similar GPA so for example the two students with the highest GPA would be matched up together and the two students with the lowest GPA would be matched up together then each pair will be split up and randomly assigned to one of the two treatments so for a pair one person will get the sleep deprived treatment and the other person will get a normal amount of sleep then we would compare the results for each pair and we will also compare all of the results so for example if we saw the the sleep deprived people were consistently getting lower test scores than the people who got a regular amount of sleep then we can say that sleep deprivation was actually the cause for a bad test score so let's quickly recap these are the experimental designs we talked about in a completely randomized design each experimental unit is randomly assigned to a random group to receive a different treatment this is what the diagram would look like group N or treatment N means you can have any number of groups groups or treatments a randomized block design first assigns people into a block based on a characteristic that is expected to influence the response of the experimental units to the treatments then a completely randomized design is performed within each block this is what the diagram would look like block N means that you can have any number of blocks that you want for a matched pairs design we saw two scenarios one that uses the same experimental units and one that uses similar experimental units both of these involve the comparison of only two treatments if you are using the same experimental units each experimental unit gets both treatments this is what the simplified diagram would look like if you are using similar experimental units we first pair up similar experimental units then each pair is split up and randomly assigned to one of two treatments this is what the simplified diagram would look like like