Presentation Transcript

1. Research Methods Unit 2

2. What is the Scientific Method? • Scientific Method: A specific set of procedures used by scientists trying to prove an idea. • Hypothesis: The specific question being asked by research, or a prediction of the outcome of a study.

3. What is the Scientific Method?

4. Types of Data • Quantitative Data: Data that is measurable using scientific methods, and that can be manipulated using mathematical equations. • Example: People who have a certain height, and a certain weight, can be calculated as having a certain BMI. • Qualitative Data: Data that is not measurable using scientific methods and cannot be manipulated using mathematical equations. • Example: Descriptions of patient behaviors.

5. Measurements • If you use any type of assessment or measurement (e.g., a test or survey) you must make sure it has two things…. • Validity: Whether the instrument measured what it was intended to measure. • Reliability: Whether the instrument will measure the same basic score across repeated testings.

6. Research Design • Research Design determines whether the study is going to be experimental or correlational. • Correlational studies show that a relationship already exists between A and B. • Experimental studies are designed to show that A causes B.

7. Correlational Design • Correlational Studies are designed to examine an existing relationship between two variables. • Researchers DO NOT manipulate any variables! • CORRELATION DOES NOT EQUAL CAUSATION!!!!! • We are only evaluating a relationship that ALREADY exists!

8. Correlational Design • Positive Correlation: The variables move together along the graph. • If one increases, the other increases.

9. Correlational Design • Negative Correlation: The variables move opposite to each other along the graph. • If one increases, the other decreases.

10. Correlational Design • Correlation is designated as an r score. • Correlations range from -1 to +1. • The closer to 1 the score is, the stronger the correlation. • R = 0.98 is strong! • The closer to 0 the score is, the weaker the correlation. • R = 0.01 is weak!

11. Correlational Design

12. Experimental Design • Groups: At least one experimental and one control. • Experimental Group: This group receives whatever treatment is being studied. • Control Group: This group goes through every step of the research study that the experimental group goes through, but does not receive the treatment. • When we evaluate data, we compare the results of the experimental and control groups. • If there is a change in the experimental group but not control, then the results are not due to chance. • If there is a change in both groups, it probably isn’t due to your treatment.

13. Experimental Design • Variables: Factors that will be measured when conducting an experiment. • Examples: Weight, gender, height, eye color, etc. • Independent Variable: The variable that is actively affected by the “treatment” in the study. • Dependent Variable: The variable being measured to determine if the independent variable had an effect on the subject. • Example: Pain control using Morphine. • Independent Variable: Amount of morphine given. • Dependent Variable: The amount of pain the patient reports after receiving pain management.

14. Experimental Design • Experimenter Bias: A condition where the researcher has affect on the outcome of the study. • Can be intentional or unintentional.

15. Research Methods • Once you decide whether you are collecting experimental data or evaluating correlational relationships, we must decide HOW we are going to look at the problem!

16. Case Studies • Case Study: An in-depth study of one or a small number of individuals. • Data is typically qualitative.

17. Observational Studies • Observational Studies: Studies that involve observing and possibly measuring behaviors or characteristics of a subject.

18. Observational Studies • Laboratory Observation: Observation that occurs in a laboratory setting under controlled conditions.

19. Observational Studies • Naturalistic Observation: Observation that occurs in a natural setting where behavior and conditions are not controlled. • Direct/Declared Observation: Those being observed are aware that they are part of a study and are being observed. • Undeclared/Indirect Observation: The researcher is not hidden, but does not allow the subjects to be aware they are being observed for the purpose of a study. • Participant Observation: The researcher integrates themselves into the lives of those being observed. • Can be declared or undeclared.

20. Surveys • Survey: A series of questions designed to learn a considerable amount about a variety of potential topics. • Personal characteristics or preferences • Life experiences • Attitudes • Opinions • Behaviors • Education level • Etc.

21. Surveys • Can be used to supplement other research data (e.g. demographic information) • Can be in a variety of formats.. • Fill-in-the-blank • Multiple choice • Written answer/Paragraph answer • Oral/Spoken Response/Interview

22. Surveys • The downside: • Can be expensive to print and mail • Long-distance phone calls get expensive • Could end up in a bad part of town doing interviews • Only around 10% of surveys end up returned • If you want 100 answers, you have to send out 1,000 surveys! • At $0.42 per stamp to send, plus a stamp to return, that makes $840 just in stamps!!!!!

23. Choosing Participants • Population: A group of individuals with certain characteristics in common. • Age • Race • Language • Religion • Geographic Location/Area • Education level • Career/Field • Confounding Factors: Something particular to a participant that may potentially interfere with your study/data interpretation.

24. Choosing Participants • Random Sampling: Choosing individuals using some sort of random method that ensures participants are not picked based on any particular traits or characteristics. • Example: Putting every potential population members’ name in a hat, then picking at random.

25. Choosing Participants • Systematic Random Sampling: Choosing the participants in some way that is random, but still has a set pattern. • Example: Every 5th name in the phone book.

26. Choosing Participants • Stratified Random Sampling: Dividing the population into sub-groups, then using a random sampling method to select individuals from each group. • Example: Dividing a population of “medical workers” into techs, medics, LPNs, RNs, Pas, Drs, etc., then selecting 10 of each.

27. Choosing Participants • Quota Sampling: Dividing the population into sub-units as in Stratified Random Sampling, then selecting a specific number from each sub-group. • Example: Dividing population into male and female groups, then selecting exactly 200 men and 300 women for data. • PROBLEM: NOT RANDOM!

28. Choosing Participants • Matched Random Sampling: Analyzing the characteristics of the population being studied, then ensuring that each known characteristic is “matched” across all groups. • Example: If a participant comes in with an IQ of 115 and is placed in Group A, the next participant with an IQ of 115 would go in Group B, and so on, until all groups had one individual with an IQ of 115. • BEST METHOD for preventing CONFOUNDING FACTORS

29. Choosing Participants • Convenience Sampling: Selecting an individual to participate in a study because they happen to be at hand or convenient to get to participate. • Example: When I do research at Austin Peay, I convince friends that teach there to offer their classes extra credit to participate. • Problem: IT’S NOT RANDOM AT ALL!

30. Length of the Study • Longitudinal Design: Follows the same group of individuals for an extended period of time. • Example: 7UP documentary series followed a group of individuals in England from the age of 7 to 47 in a series of filmed interviews. • Advantages: • You can determine if characteristics are stable across time - Developmental trends can be observed • Disadvantages: Attrition, cost, multiple researchers might be needed. • Attrition: The process of participants leaving a study before completion. • Death, illness, loss of interest, inconvenience, moved away, etc.

31. Length of Study • Cross-Sectional: Attempts to study a particular trait relatively quickly, so matches participants with similar characteristics into multiple age groups. • Example: The 7UP series could have been filmed all at once with a separate group of individuals for each age range. • E.g. a group of 7-14, a group of 15-25, a group of 25-35, a group of 35-47, etc. • Advantage: Much faster research time! • Disadvantages: Possible cohort effects. • Example: There might be something special about individuals who were alive during WWII, compared to those at a younger age.

32. Length of Study • Sequential Research: A blend of longitudinal and cross-sectional design. • Follows multiple groups for a length of time sufficient to tell if the results are due to a confounding factor or cohort effect. • Example: Instead of groups of 10-20, 20-30, and 30-40 studied for one year, each group could be studied for 20 years. • We would see longitudinal data of the 10 year olds becoming the 30 year old group by the end. • Advantages: Longitudinal developmental data available at a much quicker and cheaper rate.

33. Ethics • Institutional Review Board (IRB): A special panel at all major universities that determine if research is allowed to occur. • Decision is based on cost-benefit analysis: Is what we can learn from the research worth the potential risk of harm to the participants?

34. Ethics • Privacy: • All data that could be potential incriminating must be kept separate from the names and demographic information of a participant so that they could not be identified if the data were published. • All results must be kept confidential and only discussed with other researchers, NEVER using identifiable information. • Example: Tornado of 1999, professors at APSU seen chasing papers across the quad to collect potentially incriminating surveys regarding drug use.

35. Ethics • Informed Consent: All research participants must sign a document stating that they understand all potential risks for participating in the study. • Example: Mrs. Mayo’s Training Studying • Our goal: Determine if someone can learn to fold an Origami penguin better if they watch someone in person, watch a video, or read paper instructions. • Austin Peay’s IRB decided we must inform them… • You could receive a paper cut, so we will have non-latex bandages and a variety of antibiotic creams available. • You could become frustrated if you can’t fold the penguin right, so we will provide assistance in finishing and access to a licensed counselor to talk to.

36. Ethics • Deception: The process of not telling the participant something regarding the study to prevent them either intentionally or unintentionally altering the results. • Deception is only used if the researcher can prove that a participant knowing the goal could potentially damage the results. • The participant must undergo debriefing immediately afterword to inform them of the deception.

37. Ethics • Examples of Deception used when testing a new medication: • Blind Studies: The patient doesn’t know if they’re receiving a placebo or the new medication. • Double-Blind Studies: Neither the patient nor the doctor know whether the new medication or a placebo are being used. • Triple-Blind Studies: The patient, the doctor, and the researcher bringing the meds to the doctor all are in the dark as to whether it is the new med or placebo.