Upgrade to remove ads
Only $35.99/year

RD 1- Observational Research Designs

Terms in this set (31)

observational research (What is observational research? What is the relationship between these designs and the 4 major types of research (descriptive, correlational, experimental, quasi-experimental)
- measures a variable through observation
- most often used in descriptive
naturalistic observation
observing people in everyday life without researcher influence, cannot be experimental, is usually conducted in the field
case study
an in-depth analysis of one particular case of a phenomenon. Usually used to study rare occurrences or conditions that one cannot inflict on people in an experimental design (ex: the case study of Phineas Gage's damage to frontal lobe to understand the function of different parts of the brain)
differences in methods (Do "naturalistic design", "observational research", "field research" and "observational research" all refer to the same types of research?)
1) Observational can be done in the field or the lab, CAN be experimental (different between naturalistic and observational)
2) Naturalistic design: watching everybody in everyday life, no researcher influence
3) Field research: just the setting that the research is being conducted
4) lab research: they come to you in a controlled lab
- All refer to similar research→ observe participants in natural state, depends on how much influence the researcher has
systematic observation methods
1) ecological methods
2) relatively structured method
3) structured method
ecological methods
take into account the setting, the behaviors you see that depend on the setting, environment has specific impact on the behavior (ex: going to the gym behaviors can be different from seeing 601 behaviors)
relatively structured method
go to 601 and write everything (difficult to compare and difficult to write everything down, observe everything)
structured method
a checklist, maybe fixed-format go to 601 and write down when certain behaviors are done
unobtrusive (What does it mean to be unobtrusive? Why is it important for observers to remain unobtrusive in observational research?)
-means that the researcher attempts to blend into the setting they are observing and try not to interfere with the setting
- reduces the chances of reactivity in participants, which is when participants respond to being watched
- important because it minimizes the researcher's effects on the setting so that the data truly reflects the state of affairs
observational study steps (What are the steps one needs to follow when conducting an observational study?)
1) Determine variables and ways of measuring them
2)Select setting and mode of observation
3)Select sampling strategy
4) Train observers
5) Collect data
unacknowledged participant
- an observational design
- the people you are observing don't know you are coding them and you participate in the social scene instead of purely observing (ex: undercover cult research)
acknowledged participant
- an observational design
- the people you are observing do know that you are coding them and you participate in the social scene instead of purely observing (ex: an outsider entering a different cultural setting as a researcher and living daily life with participants)
acknowledged observer
- an observational design
- the people you are observing do know that you are coding them and you do not participate in the social scene, instead trying to remain unobtrusive (ex: an acknowledged analyst coming into a company to assess worker morale)
unacknowledged observer
- an observational design
- the people you are observing don't know that you are coding them and you do not participate in the social scene, instead trying to remain unobtrusive (ex: our dining hall behaviors study)
demand characteristics
When participants think they know what researchers are looking for, so they change behavior
event sampling
- one of the 3 types of sampling
- because you can't sample everyone at every event, can try to do as much at a single event
time sampling
- one of the 3 types of sampling
- as much at a certain time
situation sampling
- one of the 3 types of sampling
- as much as possible but in preselected situations (our study)
coding system considerations (What kinds of considerations are important when creating a coding system?)
1) which systematic observation method to use based on what best fits their variable of interest and practical constraints
- ex: relatively unstructured method would be realistically difficult to accomplish in a dining hall setting because there is too much to code
2) types of observational designs must be considered and one must choose if they should/can be acknowledged/unacknowledged or a participant/observer
3) Lastly, the sampling method must be chosen based on practical constraints and the focus of the study.
observer bias
- the bias that the researcher is subject to when coding
- expectancy effects and observer drift
expectancy effects
- an observer bias
- when the observer expects a given result which unconsciously influences their coding and interpretation of events
observer drift
- an observer bias
- when the researcher's coding gets less reliable as time passes in one session of coding due to fatigue
reactivity
- When a participant knows they are being watched and their behavior is changed
- Ex: A child knows there is a camera, and will act differently than they would
- Can reduce this by having camera rolling while explaining procedures, so they might forget it's there
habituation
- Decreases reactivity, because behavior returns to baseline
- Stop performing, go back to what they were doing before they knew they were being watching (desensitization)
- Participants will forget that they are being observed
reactance
- When people know that the researcher is taking away free will/choices, so they believe their free will is being restricted
- Ex: 3 year old having tantrum, know they can't do it but do it anyways
- May respond the opposite way, because you aren't giving them enough choice to do what they would want to do
- Threat to validity: may get annoyed and screw with data
- Participants may rebel against your study
confounds
- additional variables affecting and changing your measured/dependent variable that you didn't account and control for before running your study
operationalizing your variable
operational confounds
- confounds in the way you measure whatever you are interested in... if your measure measures more than what you are interested in, it confuses what exactly you are looking at
- can be dealt with by improving your measure through increasing reliability and better
person and environmental confounds
1) individual differences/variables that come with the person, that also relate to/vary with the predictor variable you are interested in, but can also influence the dependent variable
2) aspects of the situation that bring out a particular behavior, thus leading to the behavior (as opposed to what you are looking at)
- can often not be predicted, but can be accounted for in the analysis of data (for both)
IRB and observational studies
- some don't need approval, such as observational research conducted in public settings that do not observe illegal behavior nor pose any risk to the participant and observational research done for class projects that is not going to published.
- research that occurs in a lab setting, focuses on illegal activity, or poses a risk to participants needs to go through IRB approval.
observational research strengths
- Lack of researcher influence on the setting being observed. This way, more accurate and natural results might be obtained.
- Observational research tends to be easier to conduct because an experimental study or survey does not need to be created.
- It is also possible to conduct observational studies without IRB approval, which makes this type of research easier and less time consuming
observational research weaknesses
1) Lack of researcher control over variables. The lack of control that the researcher has on the setting of an observational research study will sometimes result in confounds or unexpected incidents during coding which could be prevented or accounted for in a lab setting. The researcher has no control over environmental variables and, to a degree, participant and dependent variables
2) Cannot assert causal relationships between variables because there is no experimental design
Sets found in the same folder