Question

Match each of the research designs below to the appropriate statistical test from the list below.

1. z Test
2. Single Sample t-Test
3. Paired Samples t-Test
4. Independent Samples t-test
5. One-way Between-Groups Analysis of Variance
6. One-way Within-Subjects Analysis of Variance
7. Two-way Analysis of Variance
8. Correlation

1. You have two nominal independent variables and one scale dependent variable. For example, you want to know if the various types of cholesterol drug are more effective at lowering cholesterol and if this effectiveness is different for men or women.
2. You are comparing a sample to a known population. You know both the population mean and standard deviation. For example, you want to know if children born to mothers who have smoked marijuana during pregnancy have children with lower than average IQ. You select a sample of 100 children whose mothers were known to have smoked marijuana during pregnancy and compare their scores to the national average intelligence test score of 100. The standard deviation for this test is known to be 15.  
3. You have one scale dependent variable, and one nominal independent variable with three levels. For example, you want find out if the effects of a new drug on bipolar symptoms depend on what type of Bipolar Disorder the patient has: Bipolar I, Bipolar II or Bipolar N.O.S. (not otherwise specified).  
4. You are comparing a sample to a known population. You know the population mean but not the population standard deviation. You want to know if the students in your district are reading at the national average according to a particular reading test. You select a sample from your district to take the test and compare it to the national average. Unfortunately, you do not know the population standard deviation for this test.
5. You want to know if children’s obesity levels is associated with exposure to food commercials on tv. For each of 100 children, you find out the child’s BMI score and the number of hours he or she spends watching tv.  
6.   You use a repeated measures design to find out if visualizing a task helps improve athletes’ performance in long-jumping. For each athlete, you ask them to jump once as they normally do (without visualizing) and once immediately after visualizing a successful jump. You measure the distance jumped on each attempt.
7. You want to know if a certain anti-psychotic drug becomes more effective over time at reducing symptoms of schizophrenia. You give schizophrenic patients the drug and measure their symptoms using the symptom rating scale after 2 months, 4 months, 6 months and 8 months.  
8. You want to know if left-handed people score higher than right-handers in a test of spatial arrangement.  

Answer 1

We are asked to match the research designs with the appropriate statistical tests

You have two nominal independent variables and one scale dependent variable. For example, you want to know if the various types of cholestrol drug are more effective at lowering cholestrol and if this effectiveness is different for men or women.

The appropriate statistical test for this research design is two-way analysis of variance because we have two independent variables.

Hence the correct option is for this research design is G.

You are comparing a sample to a known population. You know both the population mean and the standard deviation. For example, you want to know if children born to mothers who have smoked marijuana during pregnancy have children with lower than average IQ. You select a sample of 100 children whose mothers were known to have smoked marijuana during pregnancy and compare their scores to the national average intelligence test score of 100. The standard deviation for this test is known to be 15.

The appropriate statistical test for this research design is z-test because of two reasons: firstly, we know both the population mean and the stabdard deviation and secondly the sample is greater than 30.

Hence the correct option for this research design is option A.

You have one scale dependent variable, and one nominal independent variable with three levels. For example, you want to find out if the effects of a new drug on bipolar systems depend on what type of Bipolar Disorder the patient has: Bipolar I, Bipolar II and Bipolar N.O.S. (not otherwise specified).

The appropriate statistical test for this research design is one-way between-groups analysis of variance because we have only one independent variable and the effects of the new drug are being compared between the three groups:- Bipolar I, Bipolar II and Bipolar N.O.S.

Hence the correct option for this research design is E.

You want to know if children's obesity levels are associated with exposure to food commercials on tv. For each of 100 children, you find out the child's BMI score and the number of hours he or she spends watching tv.

The appropriate statistical test for this research design is correlation because correlation tells us the how strong or weak the association or relatonship is between two variables.

Hence the correct option for this research design is H.

You use a repeated measures desgn to find out if visualizing a task improves athletes' performance in long-jumping. For each athlete, you ask them to jump once as they normally do(without visualizing) and once immediately after visualizing a successful jump. You measure the distance jumped on each attempt.

The appropriate statistical test for this research design is one-way within-subjects design analysis of variance because in within-subjects analysis of variance repeated measures(asking to jump twice in our case) are taken on each subject(each athlete was asked to jump twice once without visualizing and once after visualizing).

Hence the correct option for this research design is F.

You want to know if a certain anti-psychotic drug becomes more effective over time at reducing symptoms of schizpphrenia. You give schizophrenic patients the drug and measure their symptoms using the symptom rating scale after 2 months, 4 months, 6 months and 8 months.

The appropriate statistical test for this research design is paired sample t-test because in our case the same patients are given the drug and their symptoms are measured using the symptom rating scale after 2 months, 4 months, 6 months and 8 months.

Hence the correct option for this research design is C.

You want to know if left-handed people score higher than right-handers in a test of spatial arrangement.

The appropriate statistical test for this research design is independent samples t-test because in our case we have two samples-left handed people and right handed people and these samples are independent of each other i.e., a person who is a part of left handed people is not a part of the right handed people.

Hence the correct option for this research design is D.

You are comparing a sample to a known population. You know the population mean but not the population standard deviation. You want to know if the students in your district are reading at the national average according to a particular reading test. You select a sample from your district to take the test and compare it to the national average. Unfortunately, you do not know the population standard deviation for this test.

The appropriate statistical test for this research design is single sample t-test because of two reasons: firstly, we only know the population mean but not the population standard deviation for this test and secondly, we are only collecting a single sample and comparing it to the national average.

Hence the correct option for this research design is option B.