Concerns about Case Studies
Concerns about case studies typically centre around three issues:
1. The difficulty of drawing clear causal conclusions,
2. The generalizability of the findings,
3. The potential for observer bias.
1. The difficulty of drawing clear causal conclusions in Case Studies
Case studies generally are not well suited to drawing clear causal conclusions. There are usually many factors that could be responsible for why a person has acted a certain way. These potential causes may operate individually or in various combinations, and in a case study, it is very difficult to sort them out. In an experiment, single factors and specific combinations can be isolated and manipulated as IV’S. Other factors that might affect behaviour, of which could provide alternate explanations for any finding, can be controlled. The effects of the IV on behaviour are then measured. In case studies, researchers simply do not have the degree of control over variables needed to rule out alternative explanations.
2. The generalizability of the findings in Case Studies
The second concern about case studies, generalizability, applies to all research, but the question is whether findings based on only a single case run an especially high risk of having low external validity. In any study, there is always a risk that the sample of people studied may be atypical, from the population from which they are drawn. This is much more likely to happen with a sample of one person rather than many people, but that doesn’t mean that the results from a particular case study won’t generalize. As with other types of studies, psychological principles suggested by case studies need to be validated by further research.
3. The potential for observer bias in Case Studies
The third concern, *observational bias*, occurs when researchers have expectations or other predispositions that distort their observations. The lateral point is analogous to how our brains can organize different perceptions of the same physical stimulus, as happens with ambiguous or reversible stimuli. Researchers expectations can also influence the behaviour of the people they are studying. In experiments and some other types of research, procedures can be used that will reduce the influence of such bias (for example, having a trained assistant observe participants). But in many case studies, the strong reliance on the researchers direct observations, and the prolonged and often intensive direct contact between the researcher and participant, typically invite a greater potential for observer bias that occurs with experiments.
Encompasses different types of non-experimental studies in which behaviour is systematically watched and recorded.
Basic Characteristics of Observational Research
Researchers measure the behaviour of multiple people or nonhuman animals, either in real time as the behaviour unfolds or upon reviewing electronic records of the behaviour.
Qualitative Approach in Observational Research
The goal is to achieve a holistic description and understanding of behaviour primarily through non statistical means.
Quantitative Approach in Observational Research
Relies heavily on numerical measurements and statistical analyses to describe and understand behaviour.
Mixed-Methods Approach in Observational Research
An approach the combines both the qualitative and quantitative approaches.
Ex. Jane Goodall’s research with Chimpanzees
Why Conduct Observational Research?
Like case studies, observational research is well suited to describing behaviour. But in contrast, it records the behaviour of multiple participants rather than just one. This allows observational researchers to examine relations among variables. Both in describing behaviour and in examining relations among naturally occurring variables, observational research can be exploratory or it can be used to test hypotheses and theories. Although it is poorly suited to drawing clear causal conclusions, observational findings may suggest possible causal relations that can be examined subsequently using controlled laboratory experiments. Observational studies also help to establish generalizability of principles previously discovered in experiments. When practical and ethical constraints makes it difficult or impossible to conduct experiments on a particular issue, observational studies often remain a viable approach for gathering information.
Types of Observational Research
Observational studies vary in the naturalness of their settings, in whether participants are aware or unaware that their behaviour is being observed, and in the degree to which the observer intervenes in the situation. Of the many types of observational studies, we examine five: *naturalistic observation* (disguised and undisguised), *participant observation* (disguised and undisguised), and *structured observation*.
Researchers passively observe behaviour in a natural setting. The word passively means that the researchers try to avoid direct involvement with the individuals who are being observed.
Disguised versus Undisguised Naturalistic Observation
Classified based on whether the individuals being studied are aware that they are being observed.
Undisguised Naturalistic Observation
That fact that participants being observed was not hidden from them for example, using video cameras to record their behaviour. Other examples occur in studies where researchers are visible while passively observing animals in the wild, or children in their homes or at school.
Disguised Naturalistic Observation
Occurs when individuals are not aware of the observers presence.
Advantages of Naturalistic Observation
It examines behaviour under ecologically said conditions. One aspect of ecological validity addresses the similarity between the research setting and settings that occur in real life. Here, the research setting is “real life”. Disguised naturalistic observation has the additional advantage that the behaviours being observed have not been distorted by the presence of the observer. Ultimately, the external validity of both observational and experimental findings needs to be established by conducting are studies in diverse settings and with different populations.
Disadvantages of Naturalistic Observation
They stem from the complexity of behaviour, the lack of control over the research setting, and the practical difficulties of observing every important behaviour that takes place. Many factors are likely to influence participants behaviour at any given time, and the observer does not have the ability to control the research setting in order to isolate some factors while holding others constant. The research will need to sample certain individuals, certain times, certain types of behaviours, or all of the above.
Naturalistic Research Ethics
Conducting observational research also involves ethical issues as under most circumstances researchers must obtain peoples voluntary informed consent prior to studying them. The APA permits naturalistic observation research if several conditions are met:
1. The study is not expected to cause participants harm or distress;
2. Confidential information is protected; and
3. If participants responses were to become known, this would not expose them to social, economic, or legal risks.
Additional ethical requirements govern disguised video and audio recording.
Reactivity Problem of Undisguised Naturalistic Observation
Undisguised naturalistic observation enhances the risk that the observers presence will cause the participants to alter their behaviour. This problem, called *reactivity*, occurs when the process of observing (or otherwise measuring) behaviour causes that behaviour to change. Even if the observer is not physically present, if participants are aware that their behaviour is being recorded, they may not act naturally. There are techniques that researchers can use to minimize reactivity.
The observer becomes part of the group or social setting being studied. It has been used to study behaviour in a great variety of settings and can be disguised or undisguised of which both run the risk that the researchers presence in the group, and likely interactions with group members, may influence the groups behaviour.
Disguised Participant Observation
They researcher becomes part of the group they are studying and they withhold from the other group members the fact that they are observing the groups behaviour for scientific purposes.
Theory of Cognitive Dissonance
The essential premise being that people have a psychological need for their cognitions to be consistent with one another. When two cognitions are inconsistent, people will experience and unpleasant psychological state of tension that Festinger called “cognitive dissonance”. This distance theory went on to become a highly influential social psychological theory; one reason for its appeal was that it made predictions that were supported by research, but that also seemed to fly in the face of everyday common sense.
Undisguised Participant Observation
Avoids the ethical issue of deception that arises when researchers conceal their identity. The tradeoff concerns whether participants explicit knowledge that this “new person among us” is a researcher will, by itself, cause participants to become more self conscious and alter their behaviour. This method is often combined with other methods of gathering data directly from participants.
A *qualitative* research approach that often combines participant observation with interviews to gain an integrative description of social groups. Its key goal is to explore the contexts in which behaviours occur and the meaning of those behaviours as perceived by group members. Results from ethnographic research typically are described in a narrative, story-like form, rather than through statistical analyses of numeric data.
Participant Observation Advantages
Its key benefit is the opportunity to study peoples behaviour from the viewpoint of an insider. Compared to naturalistic observation, participant observation provides a greater opportunity to gain insights about the personal meaning of behaviour to group members, particularly when the observation is undisguised and the researcher is thus free to supplement observations with interviews or other data collection methods. As for disguised participant observation, in some situations it may be the only way for psychologists to gain access to a group.
Participant Observation Disadvantages
Compared to naturalistic observers, participant observers (disguised or undisguised) run a greater risk of influencing the behaviour they are studying because their involvement in the situation is more active. The fact that participant observers are interacting with group members, and doing so for period that may last for months or even a few years, also raises the risk of losing objectivity and can make it difficult to maintain appropriate boundaries with group members.
A researcher fully or partially configures the settings in which behaviour will be observed. The researcher may present participants with specific tasks, or expose them to a social situation that the researcher has created, and observe their responses. These tasks and situations are often designed to be analogous representations of circumstances that occur naturally. In such cases, this type of structured observation is sometimes called *Analogue Behavioural Observation*.
Advantages of Structured Observation
In general, compared to naturalistic and participant observation, there is greater efficiency and control. By structuring the observational setting, the researcher can expose participants to the same tasks, make things happen at a specific time and place, and save time and money by eliminating potentially long waits for the target behaviours to occur spontaneously. Also, if the setting is a laboratory, this exposes all participants to the same physical environment and eliminates sources of distraction that might interrupt the observation process or affect participants behaviour.
Disadvantages of Structured Observation
Arise from the fact that the setting, especially in the laboratory, is only an analogue of real life. This raises the questions of whether the patterns of behaviour and the association between variables that are observed under structured circumstances represent what actually occurs under natural circumstances. It is rarely disguised, which increases the potential for reactivity to occur.
Observations of behaviour can be recorded in many ways. A researcher may use one approach or combine several. In observational research, detailed descriptions of behaviour typically come from narrative records or field notes. *Qualitative analyses* often include quotes from participants that illustrate key themes.
Used in observational research where they provide an ongoing description of behaviour that is used for later analysis. Observer writes down or narrates and extensive, continuous description of participants behaviour as it unfolds. Also come in the form of video and audio recording of participants behaviour during observation sessions. They are used for *both quantitative and qualitative analyses*.
In general, they are less comprehensive than narrative records; observers use them to record important impressions or instances of behaviour. In some cases, field notes are written soon after observation is finished rather than while it is in progress.
Behavioural Coding Systems
Involve classifying participants responses into mutually exclusive categories, and are a major component of observational studies. Observer may use a behavioural checklist to mark down the viewed behaviour into a category. It takes a great deal if work to develop a reliable and valid coding system. The coding categories must be clear and mutually exclusive, and the observers well trained, so that different observers watching the same behaviour will assign it the same codes. Coding systems exist for numerous types of behaviours. A coding system and the observers who use it constitute a measurement device.
Observer Rating and Ranking Scales
Used to evaluate participants behaviour or other characteristics. In observational research, the primary observers are typically the researchers themselves or their highly trained assistants. Acting as self-observers, participants may be asked to rank themselves, or to keep diaries of their behaviour. Studies that collect data solely through participant diaries, or through ratings made by people who know that participants, are typically *NOT considered observational research*. Such research might be called a *diary study*, a *questionnaire study*, or a *correlational study*. It is the presence of trained observers and a focus on recording ongoing behaviour that makes research observational. Nevertheless, rating and ranking sales, peer nominations, diaries, and other forms of data collection (e.g., interviews) may be used to supplement ongoing behavioural observations.
Ask members of a group to name a few members who best meet some criterion. They can be in the form of a ranking scale, although they don’t have to be.
A diary asks participants to record their behaviours or experiences for defined periods of time or whenever certain event s take place. They may contain general and specific questions that call for unstructured responses, they may include a checklist to record the frequency of certain behaviours, and they may even contain rating scales.
Before conducting a study, researchers ask themselves some basic questions and the answers to the questions will necessitate different types of sampling, with the ultimate going being to obtain a set of representative data.
Used to select a particular member (or unit, such as parent-infant dyad) who will be observed at any given time. With enough time relative to the size of the group, each member may be observed several times, for the same or similar total amount of time. The sequence of members selected could be determined randomly or chosen on some other basis.
At preselected times, the observer rapidly scans each member of a group, so that the entire group is observed within a relatively short period. Multiple scans of the entire group are taken throughout the study. The preselected times may be hours or only minutes apart. As the members are scanned, the observer records instances in which a particular behaviour occurs. To be effective, the behavioural coding system should only have a small number of categories so that the researcher can rapidly shift observations and data recording from one member to another.
Used to establish diverse settings in which behaviour is observed, which in turn *increases the external validity* of the findings compared to sampling behaviour in only on setting.
When researchers are unable to record behaviour continuously throughout the entire period of a study, *time sampling* is used to select a representative set of time periods during which observation will occur.
Concerns About Observational Research
1. Causal Inference
2. Observer bias
3. Inter-observer reliability
Problems with Drawing Causal Conclusions in Observational Research
Many variables may simultaneously affect participants behaviour and the researcher does not have to capability to isolate one or two key factors while controlling for the rest. A researcher may, however, anticipate some possible alternative explanations for the expected findings and gather data to rule them out.
Observer Bias in Observational Research
*Observer bias* occurs when the researchers have expectations or the predispositions that distort their observations and can occur in any type of research in which observers record behaviour. Expectations can shape how our brain organizes and interprets information.
Whenever possible, data should be gathered using *blind observation*: Observers should be kept unaware of all hypotheses being tested and any key information about participants that relates to those hypotheses.
Also called *interrater reliability*, which represents the degree to which independent observers show agreement in their observations. Various statistics are used to calculate it, depending on the type of data being recorded and the number of observers recording the same event. Obtaining high interobserver reliability requires a well developed coding system (or other measure of behaviour) and extensive observer training. No matter how good the coding system, it will be worthless if the observers aren’t trained properly.
One of the most common statistics for calculating the reliability between two observers and represents the percentage of times the two observers agree, over and beyond the degree of agreement that would be expected to occur by chance.
Reactivity and Observational Research
Reactivity occurs when the process of observing (or measuring) behaviour causes that behaviour to change. Sensing potential threat from a human observer, an animal in the wild will stop feeding, maintain vigilance, move away, or make aggressive gestures. It is undesirable because the behaviour being recorded no longer represents how people truly behave when then are not being observed or monitored. It is a concern of all types of behavioural research.
An approach that minimizes reactivity when disguised observation cannot be used. It is a decrease in the strength of a response, over time, to a repeated stimulus. Although generally reliable, it is not perfect as it may take a long time for participants to habituate, and once habituated, instances may still occur where the subject reacts to the presence of an observer.
Assesses behavioural without making people aware that behaviour is being measured. It is used to reduce the problem of reactivity. In other forms of this measure, people may be aware that they re participating in a research study, but remain unaware of the actual behaviours being measured and/or of the times of which their responses are measured.
Physical Trace Measures
Unobtrusively examine traces of behaviour that people create or leave behind. We can indirectly measure peoples eating behaviour or alcohol consumption by sifting through their garbage and recyclables. The products that people create are another type of physical trace measure.
Another possibility for measuring behaviour unobtrusively and are previously existing documents or other data that were produced independently of the current research. They can include collections of government documents, corporate files, personal letters, and many other kinds of records.
Unobtrusive measures and Archival records Concerns
They both carry their share of concerns and limitations. Disguised observation, any withholding of information or active deception that keeps participants unaware of the true behaviours being measured, and any use of archival records must be ethically permissible. Although a researchers use of physical trace or archival measures may be unobtrusive, the original data collection may not have been. When examining some topics, researchers may be able to find creative ways to measure behaviour inconspicuously, and such measures are best used in conjunction with other measurement techniques. Within a study, if different measures yield similar conclusions, the researcher can be more confident in the validity of the findings.