Identify the consequences (i. e., increase, decrease, no change) that the following procedures are likely to have on bias and sampling error in an experimental study. Assigning treatments to subjects alphabetically Increase sample size Calculating power Applying every treatment to every experimental unit in random order Using a sample of convenience instead of a random sample Testing only one treatment group without a control Using a balanced design Informing the human participants of which treatment they will receive

A) Bias: Increase

Sampling error: No change

B) Bias: No change

Sampling error: Increase

C) Bias: No change

Sampling error: No change

D) Bias: Increase

Sampling error: No change

E) Bias: Increase

Sampling error: No change

F) Bias: Increase

Sampling error: No change

G) Bias: No change

Sampling error: Decrease

H) Bias: Increase

Sampling error: No change

Step-by-step explanation:

A) Assigning treatments to subjects alphabetically: In this method, treatments are not assigned randomly but alphabetically and thus, it leads to an increase in bias because it's possible that one letter may be favoured over the other. Whereas, sampling error will not change because there's no effect on the sample size.

Thus;

Bias: Increase

Sampling error: No change

B) Increase sample size: In this method, it's only the probability of sampling error that will be increased because the more the sample, the higher the room for error.

Thus;

Bias: No change

Sampling error: Increase.

C) Power is the probability of rejecting the null hypothesis when the alternative hypothesis is true.

Now, this has no effect on the bias or sampling error.

Bias: No change

Sampling error: No change

D) Applying every treatment to every experimental unit in random order: In this there is possibility of an increase in bias because it's possible that some units may or may not have as much treatment as others. But it doesn't affect sampling. Thus;

Bias: Increase

Sampling error: No change

E) Using a sample of convenience instead of a random sample: In this method, there is a high risk of an increase in bias because the sample was selected conveniently instead of randomly. Thus;

Bias: Increase

Sampling error: No change

F) Testing only one treatment group, without a control group; In this method, since there is no control group, there is high probability of an increase in bias but it doesn't affect sampling error.

Thus;

Bias: Increase

Sampling error: No change

G) Using a balanced design; In this method, it means we will not have any bias as the result will be very fair, however it means there would be less of sampling error due to the fair nature of the design. Thus;

Bias: No change

Sampling error: Decrease

H) Informing the human participants of which treatment they will receive; In this method, since the human participants are informed of which treatment they would receive they could alter their responses which would lead to an increase in the bias. Whereas, it would have ko effect on the sampling error. Thus;

Bias: Increase

Sampling error: No change