Threats to External Validity.

External Validity refers to how far the results of a study can be generalized  (applied to the rest of the population). Any characteristics of the study that limits generalization is a threat to external validity. The threats are usually grouped into 3 major categories, and in this post I will discuss category 1.

In Category 1, the results of a study are demonstrated with a particular group of individuals. Selection Bias can lead to a biased sample of results, which is a major threat to external validity. Selection Bias means that the sampling procedure favors the selection of some individuals over others. Another threat is the use of University or College students because they are so easily available. Sears (1986), demonstrated that college students have a stronger tendency to comply with authority, have less stable peer relationships and also higher intelligence than non college adults. This means that college students are a highly select group, and would be difficult to generalize.

Most participants of a study have signed up because they Volunteered, which brings around the next threat, volunteer bias. Volunteers are not perfectly representative of the general population. Another threat to external validity occurs whenever a study uses participants who share similar characteristics. Demographic characteristics such as age, race and gender and also socioeconomic status, can limit generalization. Finally, cross-species generalization is a threat to external validity, as research is conducted with nonhumans and presumed to be applicable to humans. Rats are a very good species to be used for research on eating  because they have similar digestive systems to humans. As a result of this, researchers can confidently generalize the results of research with rats, to humans.

Comments for TA 14/3/12

http://laurenpsychology.wordpress.com/2012/03/11/is-single-case-design-a-useful-tool-or-should-group-design-be-the-preferred-method-for-studying-psychological-variables/#comment-91

http://psuc98.wordpress.com/2012/03/11/gender-bias/#comment-50

http://alhoward.wordpress.com/2012/03/05/the-internet-and-the-death-of-ethics/#comment-63

http://lmr92.wordpress.com/2012/03/11/is-it-possible-to-operationalize-every-variable/#comment-47

The main differences between a case study and single case designs.

My understanding of a Case Study is a detailed description of one individual, which is rich in qualitative data, and gives a greater understanding of a particular disorder an individual has, or an insight into something which they went through. It is typically presented as a report and will usually contain a detailed description of the observations and experiences during diagnosis. A case study may also involve an intervention or treatment administered by the researcher, and if it does not it is sometimes called a case history. A strength of a case study, is the intense detail that is included, however a weakness is that it lacks external validity because it can only be applied to the individual and no one else in the greater population. The most famous example of a case study, is Freud’s patient Anna O, who was diagnosed with hysteria. Her symptoms and behavior were reported in the case study and helped to give a greater understanding of psychological illness at that time. 

  A single-subject design or single-case designs, are research designs that use the results from a single participant or subject to establish the existence of cause and effect relationships. Most single subjects were developed by behaviorists examining operant conditioning. The behavior of a single subject was observed and any changes in the behavior was noted while the researcher manipulated the stimulus. A particular type of design is the ABAB design (or reversal design) which is a single subject design consisting of four phases: a baseline phase, a treatment phase, a return to baseline phase and a second treatment phase. The goal is to demonstrate that the treatment causes changes in the participant’s behavior. 

  It can be seen that there are a number of differences between a case study and a single case design, but both are very useful in data collection, and showing a cause and effect relationship. 

 

 

http://en.wikipedia.org/wiki/Anna_O.

http://en.wikipedia.org/wiki/Case_study

Research Methods for the Behavioral Sciences , Frederick J Gravetter & Lori-Ann B. Forzano. 

 

 

 

 

 

 

 

comments for TA 21/2/12

http://laurenpsychology.wordpress.com/2012/02/19/is-it-ethically-okay-to-use-internet-sources-for-qualitative-studies/#comments

http://jessica0703.wordpress.com/2012/02/19/how-effective-is-the-use-of-observation-as-a-method-of-research/#comments

http://psuc98.wordpress.com/2012/02/17/can-sadness-be-operationalised/#comments

http://raw2392.wordpress.com/2012/02/17/should-milgrams-study-of-ever-taken-place-blog-for-190212/#comments

“How can research methods help us understand cause and effect relationships between variables?”

In order to understand cause and effect relationships between variables, we must carry out scientific research and to do this we need research methods. Without research methods, an experimental design would not be created and therefore no research would be carried out.

Once the independent and dependent variable have been decided, the experimental design is chosen and will either be within groups or between groups. This is how the participants will be tested, and many things have to be taken into consideration, such as time and fatigue effects in participants. Once the design has been chosen, participants need to be gathered and there are a number of ways to do this. The most common way is usually random selection of participants or advertising the need for participants using posters or the internet.

The biggest part of research methods, is research ethics. Without ethical approval, the research would not be carried out in the first place, so this is the main way in which research methods help cause and effect relationships. Research ethics include things such as Informed Consent, Debriefing, Privacy and the right to withdraw from the study and to withhold information. Participants must give informed consent before doing the study and be fully debriefed afterwards.

By using research methods and following research ethics, we can see and understand how an experiment is designed and conducted, to give the cause and effect relationship between the two variables being investigated.

Comments for TA :) 7/2/12

http://laurenpsychology.wordpress.com/2012/02/05/is-there-anything-that-cant-be-measured-by-psychologists/#comment-57

http://psuc98.wordpress.com/2012/02/04/does-correlation-show-causality/#comment-29

http://raw2392.wordpress.com/2012/02/01/blog-for-050212/#comment-40

http://psyalo.wordpress.com/2012/02/05/are-psychology-participants-too-w-e-i-r-d/#comment-61

 

Comments for Julie, Week 11

http://statsjamps.wordpress.com/2011/11/25/qualitative-research-isnt-as-scientific-as-quantitative-methods/#comment-23

 

https://psuc39.wordpress.com/2011/12/09/the-story-of-ethics/#comment-19

 

http://psud56.wordpress.com/2011/12/09/correlation-versus-regression/#comment-16

 

http://khpsud7d.wordpress.com/2011/11/15/qualitative-data-vs-quantitative-data-and-stuff/#comment-20

“How do you know whether your findings are valid?”

If a measurement procedure in an experiment is valid, then it will measure what the variable claims to measure, to a certain degree. In order for there to be full validity, the measurement process must completely measure the variable it claims to measure. In other words, whatever you say will be done in the methods, must be done.

There are many different types of validity, the most common and simple one is perhaps Face validity. Face validity is demonstrated when a measure superficially appears to measure what it claims to measure. This also makes it an unscientific form of validity as well as a simple form. Another common one is Concurrent validity, which is demonstrated when scores obtained from a new measure are directly related to scores obtained from an established measure of the same variable.  An example of Concurrent validity is IQ tests. If you had developed a new IQ test, you could demonstrate that your test really measures intelligence by showing that the scores from your test differentiate individuals in the same way as scores from a standardized IQ test.

If your measurement procedure matches up to the many different types of validity, and measures what you intended it to measure, then your findings will be valid. However this does not mean they will be reliable, because although validity and reliability go together, they are both concerned with different aspects of the research.

Comments for Julie.

http://statsjamps.wordpress.com/2011/10/28/is-it-dishonest-to-remove-outliers/#comments

http://chocolateraisons.wordpress.com/2011/10/28/what-role-does-the-null-hypothesis-really-play-in-the-scientific-process/#comment-32

http://camilia92.wordpress.com/2011/10/26/23/#comment-20

http://psuc0f.wordpress.com/2011/11/15/qualitative-research-isnt-as-scientific-as-quantitative-methods/#comment-25

‘Qualitative research isn’t as scientific as Quantitative methods’

Qualitative research, is research gathered from questionnaires or journals to give a better understanding of human behavior without using numbers. Quantitative however, is the use of numbers and data to prove a hypothesis or research question.

A popular and common example of Qualitative research is a case study, which is specifically about one person, and describes them in full and specific detail by writing up accounts of their past and how they behave everyday now. An example of a famous case study used to explain and show behavior, is Anna O who was treated by Freud for having delusional thoughts and being unable to walk or speak. One of the methods used for collecting Qualitative data, is Ethnography, where data is collected from different cultures and described for developing a theory. It can be said that Qualitative research isn’t as scientific, because thoughts and feelings can be very varied and cannot be applied to one person whereas numbers and figures will always be the same and mean the same. Also numbers are much easier to interpret than the behavior and emotions of a particular person.

There are a number of ways to collect Quantitative research, as it is the most widely known and used data type. Some ways include experimental methods, field methods and also observational methods. The data that is collected is always in a numerical format and can be correctly and scientifically analysed using statistic measures such as ANOVAs.

To conclude, I would agree with the above statement, because it is far more scientific and applicable to use Quantitative methods rather than Qualitative.