Designing a research study/Self-check assessment

From WikiEducator
Jump to: navigation, search

Use the following quiz questions to check your understanding of simple linear regression. Note that as soon as you have indicated your response, the question is scored and feedback is provided. As feedback is provided for each option, you may find it useful to try all of the responses (both correct and incorrect) to read the feedback, as a way to better understand the concept.

Icon qmark.gif
Introduction

Identify the type of study design in the following scenario.

  • Researchers want to examine the effectiveness of three programs on the weight loss of men and women in the 40-50 year old age range. 150 men and 150 women participate in the study. Subjects are randomly assigned to the three programs. They spend 3-4 hours in the program per week and they continue the program for six months. Their weight is recorded before and after the program.[1]
    • observational study
      • That's not quite right. Note that in this study design 3 different treatment programs were imposed on 3 groups of subjects. Which study design includes this feature. Try again.
    • experimental study
      • That's correct. This study design is an experiment because the researchers imposed treatments on the subjects in the study.
    • sample survey
      • That's not quite right. Note that in this study design 3 different treatment programs were imposed on 3 groups of subjects. Which study design includes this feature? Try again.
    • combination of experimental and observational study
      • That's not quite right. Either a study design is experimental in nature or observational. A combination is not possible. Try again.
    • there is not enough information to determine the study design
      • That's not quite right. The scenario indicates that researchers randomly assign subjects to three different programs (treatments), and required to participate in the program for a period of time. In which study design is the treatment imposed on the study's subjects? Try again.
  • In a large midwestern university with 30 different departments, the university is considering eliminating standardized test scores from their admission requirements. The university wants to find out whether the students agree with this plan. They decide to randomly select 100 students from each department, send them a survey, and follow up with a phone call if they do not return the survey within a week.[2]
    • observational study
      • That's not quite right. Think about which study design typically has researchers asking the subjects for their opinions. Try again.
    • experimental study
      • That's not quite right. Think about which study design typically has researchers asking the subjects for their opinions. Try again.
    • sample survey
      • That's correct. A sample survey is a study in which the subjects report the values of the variables (often their opinions on a set of questions).
    • combination of experimental and observational study
      • That's not quite right. Either a study design is experimental in nature or observational. A combination is not possible. Try again.
    • there is not enough information to determine the study design
      • That's not quite right. Think about which study design typically has researchers asking the subjects for their opinions. Try again.
  • An office manager wonders whether there is any relationship between drinking coffee before 10 am and alertness. He selects at random 3 days of the week, and in those days, he compared the alertness level of 25 employees who usually drink coffee before 10 am and 25 employees who do not usually drink coffee before 10 am.[3]
    • observational study
      • That's correct. An observational study involves recording variables' values without any attempt to influence the responses.
    • experimental study
      • That's not quite right. Which study design involves recording results without any attempt to influence the data? Try again.
    • sample survey
      • That's not quite right. Which study design involves recording results without any attempt to influence the data? Try again.
    • combination of experimental and observational study
      • That's not quite right. Either a study design is experimental in nature or observational. A combination is not possible. Try again.
    • there is not enough information to determine the study design
      • That's not quite right. Which study design involves recording results without any attempt to influence the data Try again.
  • A study indicated that elderly people (age 70 and higher) who had pets lived longer and became less depressed than elderly people who did not have pets. The data came from the records of 700 elderly people who went to a local clinic for treatment. 379 of the patients had pets and 321 did not.[4]
    • observational study
      • That's correct. An observational study involves recording variables' values without any attempt to influence the responses.
    • experimental study
      • That's not quite right. Which study design involves recording results without any attempt to influence the data? Try again.
    • sample survey
      • That's not quite right. Which study design involves recording results without any attempt to influence the data? Try again.
    • combination of experimental and observational study
      • That's not quite right. Either a study design is experimental in nature or observational. A combination is not possible. Try again.
    • there is not enough information to determine the study design
      • That's not quite right. Which study design involves recording results without any attempt to influence the data Try again.



Icon qmark.gif
What can we conclude?
  • An examination of the medical records of more than 250,000 women in the 20-40 year age range indicated that those who were overweight had longer than average labor when their first child was born. The study concluded that above average weight causes women in the 20-40 year age range to have longer labor when delivering their first child. Which of the following statements about the conclusion reached is correct.[5]
    • In order to decide whether or not this conclusion is appropriate, we need to have the scatterplot and the coefficient of correlation for the weight of the mothers and the hours of labor.
      • That's not quite right. The statistics reported in the study will not affect whether or not the conclusion can be a causative statement. Recall that causation is not a valid conclusion when no attempt has been made to impose a treatment on the subjects. Try again.
    • Given the large sample size used in the study they reached the right conclusion and they can generalize this to the overall population.
      • That's not quite right. The sample size does not affect whether or not the conclusion can be a causative statement. Recall that causation is not a valid conclusion when no attempt has been made to impose a treatment on the subjects. Try again.
    • This conclusion is not correct because there are many other factors other than weight that could contribute to long labor.
      • That's correct. The study described is observational, and makes no effort to control for potentially lurking and confounding variables.
    • They cannot draw such a conclusion because they did not randomly assign the subjects to the control and the experimental group.
      • That's not quite right. There are no experimental and control groups in the study, and it would not be possible to create such groups.



Icon qmark.gif
Designing experiments
  • Fifty subjects are each tested in both a control condition and an experimental condition. This is an example of[6]:
    • a completely randomized design
      • That's not quite right. In a completely randomized design experimental units are randomly assigned to only one of the groups. In which design might subjects be tested in each of the treatment conditions? Try again.
    • a matched pairs design
      • That's correct. In a matched pairs design, the same subjects may be tested in each of the treatment conditions.
    • a block design
      • That's not quite right. In a block design experimental units are randomly assigned to only one of the groups. In which design might subjects be tested in each of the treatment conditions? Try again.
  • Subjects are randomly assigned to either a drug condition or a placebo condition. This is an example of[7]:
    • a completely randomized design
      • That's correct. In a completely randomized design experimental units are randomly assigned to only one of the treatment conditions.
    • a matched pairs design
      • That's not quite right. In a matched pairs design, each subject (or matched pair of subjects) is tested in each of the treatment conditions. In which design are subjects randomly assigned to one treatment condition, without concern for other variables? Try again.
    • a block design
      • That's not quite right. In a block design, subjects are first grouped according to a blocking variable and then randomly assigned to treatment conditions. In which design are subjects randomly assigned to one treatment condition, without concern for other variables? Try again.
  • Separately, twenty-six male subjects and twenty-six female subjects are randomly assigned to either a drug condition or a placebo condition. This is an example of:
    • a completely randomized design
      • That's not quite right. In a completely randomized design experimental units are randomly assigned to only one of the treatment conditions, without regard for other variables. In which design are subjects first grouped according to a potential lurking variable, and then randomly assigned within group to the treatment conditions? Try again.
    • a matched pairs design
      • That's not quite right. In a matched pairs design, each subject (or matched pair of subjects) is tested in each of the treatment conditions. In which design are subjects first grouped according to a potential lurking variable, and then randomly assigned within group to the treatment conditions? Try again.
    • a block design
      • That's correct. In a block design, subjects are first grouped according to a blocking variable (typically a potential lurking variable) and then randomly assigned within group to the treatment conditions.
  • A psychologist is examining the effect of showing pictures on learning of words by seven-year-olds. The seven-year-olds are randomly assigned to two groups. The experimental group is shown the word along with the picture. The control group is shown only the word. At the end of the experiment, the subjects are given a test on the number of words they can read.[8] This experiment is not double-blind because:
    • the children were allowed to see the experimenter.
      • That's not quite right. Recall that in a double-blind experiment neither the experimenter nor the subjects are aware of which subject receives which treatment. Try again.
    • the children and experimenters are aware of which child is being administered the treatment.
      • That's correct. In a double-blind experiment neither the experimenter nor the subjects are aware of which subject receives which treatment.
    • The children were not given enough time to learn the words.
      • That's not quite right. Recall that in a double-blind experiment neither the experimenter nor the subjects are aware of which subject receives which treatment. Try again.
    • the children are randomized into two groups.
      • That's not quite right. Recall that in a double-blind experiment neither the experimenter nor the subjects are aware of which subject receives which treatment. Try again.



Icon qmark.gif
What can we conclude?

Suppose two researchers wanted to determine if aspirin reduced the chance of a heart attack. Researcher 1 studied the medical records of 500 patients. For each patient, he recorded whether the person took aspirin every day and if the person had ever had a heart attack. Then he reported the percentage of heart attacks for the patients who took aspirin every day and for those who did not take aspirin every day.

Researcher 2 also studied 500 people. He randomly assigned half of the patients to take aspirin every day and the other half to take a placebo everyday. After a certain length of time, he reported the percentage of heart attacks for the patients who took aspirin every day and for those who did not take aspirin every day. Suppose that both researchers found that there is a statistically significant difference in the heart attack rates for the aspirin users and the non-aspirin users and that aspirin users had a lower rate of heart attacks.

  • Can both researchers conclude that aspirin caused the reduction? Why? Researcher 2[9]
    • Only Researcher 2 can conclude that aspirin caused the lower rate of heart attacks, because 1) the randomization of patients to groups served to control the potential lurking and confounding variables and 2) the treatment was imposed on each group, with the control group receiving a placebo. The study designed by Researcher 2 is experimental, while the study designed by Researcher 1 is observational.



Icon qmark.gif
Designing experiments

A high school is interested to examine the overall effectiveness of its online ACT prep course, along with the effectiveness of different instructional focuses. First, the school wanted to see whether the online course was better than its standard classroom-based course. Second, the school wanted to know whether students did better on the ACT when they completed 1, 2, or 3 full-practice tests during the course. The school randomly assigned 236 students to one of the different variants of the course (i.e., classroom or online, coupled with one of the practice test options). The success of each variant was measured by the students' combined ACT score.

  • This study is an experimental (experimental/observational) study.
  • The explanatory variable(s) in this study is: xx
    • There are two explanatory variables: type of course (online vs classroom-based) and number of full-practice tests completed (1, 2 or 3)
  • The response variable in this study is: xx
    • The response variable is the combined ACT score.
  • The number of treatment conditions in this study is 6.



Matching

Hospital floors are usually covered by bare tiles. Carpets would cut down on noise but might be more likely to harbor germs. To study this possibility, investigators randomly assigned 8 of 16 available hospital rooms to have carpet installed. The others were left bare. Later, air from each room was pumped over a dish of agar. The dish was incubated for a fixed period, and the number of bacteria colonies were counted.[10]

Match the terms in column A with the study elements in column B.

A B
experimental units the 8 rooms with bare "uncarpeted" floors
treatment group the 16 hospital rooms
control group the number of bacteria colonies in a dish
response the 8 carpeted rooms

Notes

  1. Question adapted from Ebook Problems EDA IntroDesign, Problem 11 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.
  2. Question adapted from Ebook Problems EDA IntroDesign, Problem 31 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.
  3. Question adapted from Ebook Problems EDA IntroDesign, Problem 19 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.
  4. Question adapted from Ebook Problems EDA IntroDesign, Problem 20 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.
  5. Question adapted from Ebook Problems EDA IntroDesign, Problem 22 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.
  6. Adapted from Experimental Designs at Online Statistics Education: An Interactive Multimedia Course of Study. Project Leader: David M. Lane, Rice University. Retrieved 14 October 2012.
  7. Adapted from Experimental Designs at Online Statistics Education: An Interactive Multimedia Course of Study. Project Leader: David M. Lane, Rice University. Retrieved 14 October 2012.
  8. Question adapted from Ebook Problems EDA IntroDesign, Problem 10 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.
  9. Question adapted from Ebook Problems EDA IntroDesign, Problem 2 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.
  10. Question adapted from Ebook Problems EDA IntroDesign, Problems 6-8 in Probability and Statistics EBook, from UCLA Statistics Online Computational Resource (SOCR), Retrieved 14 October 2012.