Ch2- Scientific Research Flashcards
What are the 5 characteristics of quality scientific research?
1) It is based on measures that are objective, valid, and reliable.
2) The research can be generalized.
3) It uses techniques that reduce bias
4) The findings are made public
5) The results can be replicated.
Define: Objective measurements.
Give an example.
The measure of an entity or behaviour that is consistent across instruments and observers.
eg. 10kg
Define: Operational definitions
Statements used to describe procedures (or operations) and specific measures that are used to record observations.
eg. Obese could be defined as “BMI over 200”
Define: Variable
Object/concept/event being measured.
Define: Reliability
“Time tested”; being able to provide consistent and stable answers across multiple observations & points in time.
Define: Validity and Ecological Validity
“Relevance”- the degree to which the procedure accurately measures what it wants to measure.
Ecological validity is being able to apply the results to the real world.
Define: Generalisability vs. Overgeneralising
Degree to which your results can be applied to more events, groups, etc.
Overgeneralising is the inappropriate generalisation to groups that do not apply.
Define: Population
The group you want to generalise about.
Define: Sample, Random Sample, Convenience Sample
Sample: A select group of population members
Random Sample: Sample with every individual in a population having an equal chance of inclusion.
Convenience Sample: Sample most readily avaliable.
Define: Replication
(Hint: in the context of studies)
Repeating a study and finding a similar outcome each time
Define:
- Researcher bias
- Subject/participant bias
- Hawthorne effect
- Demand characteristics
- Social desirability
- Researcher bias: Bias introduced by researcher
- Subject/participant bias: Bias introduced by the subject
- Hawthorne effect: Bias introduced by the effect of observation
- Demand characteristics: Inadvertant cues given off by the researcher or experiment context about how the participants are expected to behave.
- Social desirability: Participants respond to be viewed favorably.
What do these all have in common?
Define:
- Anonymity
- Confidentiality
- Single-blind study
- Double blind study
All are techniques that reduce bias.
- Anonymity: Responses are recorded without ID
- Confidentiality: Test results will only be seen by researcher
- Single-blind study: Participants don’t know the purpose of the study, eliminating subject bias.
- Double blind study: Participants and experimenters both don’t know the exact treatment for any individual.
What are 5 Elements of poor research? Finish the statements below.
\_\_\_\_\_Hypothesis \_\_\_\_\_dotes \_\_\_\_\_selection of available data Appeals to \_\_\_\_\_\_ Appeals to \_\_\_\_\_\_
Untestable hypothesis Anecdotes Biased selection of available data Appeals to authority Appeals to common sense
Define: Research question vs. research design
- Question kickstarts research.
- Design refers to methods that allow the hypothesis to be tested.
Define: Descriptive research
Answers the question of WHAT a phenomenon is.
Define: Naturalistic observation
Unobtrusive observation of a subject in a natural environment.
Define: Correlational research, Correlation, and the Third variable problem
Correlational research: Measurement of the degree of association between two or more variables. Can be visualised with a scatterplot.
Correlation: relation between two variables. Does not imply causation.
Third Variable Problem: The possibility that a well-established correlation between 2 variables is caused by a 3rd, unmeasured variable.
Which coefficient (value) shows the strongest correlation?
a) +0.2
b) -0.5
c) -0.9
d) 1.5
c. Correlation values only go to +1.0 or -1.0.
Define:
- independent variable
- dependent variable
- confounding variable
- independent variable: The variable that is manipulated between 2 or more groups
- dependent variable: The variable that is recorded and measured.
- confounding variable: a variable outside the control of the researcher that might affect or provide alternative explanations for the results.
Define: Between-subjects design vs. Within subjects design. Which design involves a control group?
Between-subjects design involves a control group.
In a between subjects design, the performance of different groups of participants are measured. There is an experimental group that receives the treatment, and a control group that doesn’t.
A within subjects design has the same participants respond to all types of stimuli/ experience all experimental conditions.
Define: Quasi- experimental research
where 2 or more groups are selected for pre-determined characteristics, rather than by random assignment.
Define: Converging operations.
(Hint: under the context of theories being tested)
When a theory holds up under dozens of tests using a variety of designs.
Describe: Research Ethics Board (REB). Who do they protect?
A committee of researches and officials at an institution responsible for protecting HUMAN participants.
Describe “Informed consent” and “deception”. Are they mutually exclusive?
Informed consent: Knowing the risks, tasks, purpose of the study, and giving consent without pressure.
Deception: Misleading or only partially informing participants of the true nature of the investigation.
Not mutually exclusive. Deception should only be short-term; the nature of the investigation should be revealed in a debriefing once the experiment is over.
Describe scientific misconduct.
Fabrication or alteration of data to fit desired results.
Match the following:
- Normal distribution
- Positive skewed distribution
- Negative skewed distribution
___/\
/\___
__/\__
Normal __/\__
Positive skewed distribution /\___
Negative skewed distribution ___/\
Mean, median, and mode:
which one corresponds to:
50th percentile?
divided average?
most commonly found?
Mean: divided average
Median: 50 percentile
Mode: Most common
Define: standard variation.
average distribution around the mean (or average)
Define: Hypothesis test vs. Statistical significance
Hypothesis test: Statistical method of evaluating whether or not differences among groups are meaningful, or could have been arrived at by chance alone.
Statistical significance: A concept that implies that t means of the groups are further apart than you would expect them to be by random chance alone.
