http://bloggenrolla.wordpress.com/2012/03/25/the-file-drawer-effect/#comment-81
http://psud56.wordpress.com/2012/03/23/animal-testing-should-we-or-shouldnt-we/#comment-42
Psychological variables can be abstract or concrete and these can vary from individual to individual and this can be due to environmental factors, personality and across time, Andy Field, (2009). These variables are measured to examine whether factors can have an impact and cause a change, Andy Field, (2009).
A single case design is when an individual or a small number of individual take part in a study and these participants undergo both the experimental condition and the control condition, Watson (2003). Whereas group design can be within group design, this is similar to a single case deign, however on a larger scale, it includes more participants, Hall (1998). A group design can look at between group design where the participants are either in the experimental or the conditioned group, http://www.une.edu.au/WebStat/unit_materials/c2_research_design/between_repeated_measure.htm .
All of these designs have advantages and disadvantages for studying psychological variables. For example if we were to measuring what impacts motivation we could use any if these designs and all would have advantages. However in my opinion within group design would be the best method of studying these variables.
A research conducted by Carlson and Tamm, (2000) looked at children with a hyperactivity disorder and reward for performance and motivation, the method of use was a within measure design. Carson and Tamm found that the motivational influences for children with the hyperactivity disorder were through self-perceptions of their own performance. These findings can then go on and be used in the real word to help teaching procedures to help increase motivation in children with a hyperactivity disorder. Therefore the findings can be generalised to those who have the hyperactivity disorder and therefore having advantages to society. It is also useful tool as it examines the same participants therefore using the same participants will demonstrate what works and what does not.
However, although the single case design also uses participants for the experimental and the control condition, it may not be as well applied to society as a within subjects might be. Another complication may be the sample size. It is suggested that a larger sample size of around 30 participants should get a god result, Lani, 2009. Therefore when using a small sample size like in a single case design, it could have its complications.
Therefore I feel that although they all have advantages and disadvantages if you are applying it to society then a group design should be used
Andy Field, (2009) Discoversing Statistics using sppss. Tell me why my evil lecture is forcing me to use satistics. 1-30.
http://www.practicalpress.net/updatenov05/SingleSubject.html waston
http://web.mst.edu/~psyworld/within_subjects.htm hall
http://0-search.proquest.com.unicat.bangor.ac.uk/docview/614356475/135A4F1F951277794C1/2?accountid=14874 Carlson and Tamm, (2000)
http://researchsamplesize.blogspot.co.uk/2009/04/sample-size-justification.html Lani
2) http://bloggenrolla.wordpress.com/2012/03/11/bias-in-sona/#comment-65
3) http://khpsud7d.wordpress.com/2012/02/19/some-ramble-about-effect-size/#comment-33
Please could you mark these ones for me,
Thank you.
According to the American Psychological Association there are five general principles that need to be followed to be accepted by ethical standards. In order to carry out a research these five principles such be taken in to consideration. So when asking whether it is ethically alight to use internet sources as a means for collecting data for qualitative studies the five principles should be taken in to consideration. But first what is a qualitative studied? A qualitative study observes the behaviour of individuals takes questionnaires and case studies by its means of collecting data and puts it into words not numbers. Therefore using internet sources would be a great insight into understand and interpreting the minds of teenagers and gamblers. But should it be used? I think not. Not when it does not take in consideration the ethical principles. If the answer is yes then wouldn’t it be like deception in quantitative studies? This is going to look at the five ethical principles and relate to examples of internet sources and why it is not right it use it.
The first ethical principle we are going to look at is the Beneficence and Nonmaleficence. This is where participant’s rights are taken in to consideration and therefore their health and wellbeing is considered. So for example if individual wrote a blog that was personal to them about certain beliefs they may have and a researcher wanted to conduct a study on the beliefs of animal research or animal fur being used for fashion purpose, then they have not taken the right of the individual in to consideration. If there is no way to asking the individual to use their information then they are violating individual privacy. Also it could be suggested that a form of deception is being used.
The next principle that needs to be considered is the fidelity and responsibility; this is where researchers must form a responsible relationship with the participant or whomever the research is conducted from. So would it be ethical just to email the blog writer and ask permission. Is this a professional relationship to form?
A third principle that needs to look at is the integrity where one of the main attributes for this principle is honestly in the science formed. Therefore conducting information without forming consent and looking at the research before and not after is this real scientific honesty? Another issue with this is that the data or text provided may be false and this is affecting the honest of science findings.
The forth principle is known as justice, again this must mean that there is justice in what is published so therefore the individual may not want their personal thoughts and feelings in a study. The final principle is the respect for people’s rights, this again fits in with the idea of the forth principle.
To conclude it would be ethically wrong not to take any of these principles in to consideration when using internet sources for data. Otherwise it may be deemed ethically inconsiderate.
http://www.apa.org/ethics/code/index.aspx?item=3
http://www.qsrinternational.com/what-is-qualitative-research.aspx
Hello, Here are my comments. Please could you mark them for me please. 🙂
Thank you
http://jessica0703.wordpress.com/2012/02/02/debriefing/#comment-59
Should data that accepts the null hypothesis be reported?
There are two types of hypothesis one is known as the alternative hypothesis and the other is known as the null hypothesis. What is the difference? The alternative hypothesis states that an effect will be present, for example there will be an difference in reading sores for those who read for pleasure at home than those who just read at school. Whereas the null hypothesis states that there will be no difference among reading levels.* Therefore accepting the null hypothesis would mean that there is no difference or relationship between the variable and may not support other findings. Therefore in a study conducted a researcher would conclude that there is no difference and the null hypothesis would be accepted. In many cases where the null hypothesis is accepted, journals are not reported; the papers are not published. This may be due to the reason of not finding an effect and therefore not wanting to publish a report. The researcher may feel that it is not important to report on significant findings as it may not have an impact on the population, But, what if this was not the case. What if reports published with non-significance did impact the population if they were published? I feel that there are may be many reasons to why significance was not found and that if it was published it could have a beneficial impact to the population. A reason might be that a type one error has occurred and therefore a false positive has developed, whereby a significance was found and was reported to accept the null hypothesis.* If reports were published then other researchers would be able to look at the procedures and methods and come to conclusions to how the study was ran, for example they might be able to look at the tightness of control groups, the individual difference or maybe even use a different design. This could be accomplished if non-significant journals were reported and they could be compared with those that have a similar alternative hypothesis and used to see what difference there are in procedure and the participants used. An effect size is used in a process called meta-analysis, this is the process in which the effect size is compared with other studies asking similar questions about the individuals.* If the effect size was found to be large in the study with non-significance then it may have beneficial outcomes for future studies. It is also suggested that secondary meta-analysis can help improved already published journals with a poor significance. ** This is another reason why non-significant data should be published. Another reason to why non-significance data should be reported is that a researcher might be looking into something new, if this was to be published then other researchers could try and developed the ideas and procedures to try and test for a significance in the future, it may also help in schooling, where a new reading, writing and spelling scheme may be introduced and a significance wasn’t found so it was not passed on, other researchers may be able to help find different methods for the initial hypothesis to be produced. To conclude, although there may be factors causing a significance finding to be non-significant or methods needing to be improved, or for the pure fact there is nothing primarily causing a non-significance I feel that non-significance should be reported because although there is not initial relationship with in influence of other, researchers might be able to find a significance in the future and find something new. It is all to do with trial and error!
- Andy Field – Discovering statistics using spss.
** – http://stat.smmu.edu.cn/history/primary.pdf
I have decided to talk about type I and type II errors mainly because I always get confused about which way round they go, and seeing as they are going in be in the exam I thought it would be a good method to help me learn type I and type II errors before the midterm.
So what are type I and type II errors? Type I and type II errors are mistakes made by the researcher when reporting the findings of a study. The errors can have adverse effects on society and might even be damaging.
Type II errors are errors made by the researcher who claims that they have found a difference when there was actually no effect. Type II error is when the researcher fails to reject a null hypothesis that is true. This is the worse out of the type of errors as it can have the worse effects on society. Type I errors do not have such adverse effect on society as these are mistakes where there is an effect by the researcher has failed to find it and has regarded it to not have an effect. In this case the researcher fails to reject a hypothesis that is false. ( Nayman, Pearson)
An example of type I error in society could be damaging. For example, on non-psychological bases if a prisoner who was guilty of murder was found to be innocent then this would affect society. The murder has got away free and might be likely to re-offend again causing an innocent death again due to the wrong decision being made. This can be similar to a type I error in psychology as the jury have found a guilty man to be innocent, this might have occurred due to the juries not going over the evidence again to find the guilty person to actually be guilty. Type II error might occur due to this reason, the researcher may not check the data again or look at it carefully causing a studying to be released saying there is an effect when there is not.
Another example of type II error is when the researcher reports that there is a significance between the treatment of a particular drug and a placebo. If the researcher was to say that the drug was to have an effect and help the individual then again it could have adverse effect on society. If the treatment drug had no impact on the individuals health it might cause depression as they might feel demotivated as a drug that was told that would help them has not in the slightest may cause negative thoughts an low self-esteem.
Type I error is when an effective is found but the researcher claims there is no effect, he has been able to find it. This may be due to the size of the effect or due to extreme outliers not being removed. Again the researcher may have failed to look thorough the data finely before presenting the results. This may not have as much of an impact on society as the type II error. However it might have more of an impact on the individual, going back to the murder example the wrong person may have been found guilty. This may have negative effects on the individual as they have been found guilty for something they did not do. This may have occurred as the individual was at the wrong place at the wrong time and the evidence for the being may have caused the particular individual to look guilty. Another example of type I error is that there is a relationship between a treatment and placebo drug the finding might tell the reporter that there is no significance when there may have been a fault with the way it was explored. Again this would be due to the researcher not checking the data again. This would not have such an impact on society as it would not have negative effects on them as they have been told the treatment does not work. Therefore not causing any physical or mental harm to individuals receiving the drug.
To conclude type I and type II errors should be avoided when reporting results, data should be checked and ensured that no mistakes have been made. Type I errors can caused bad effects to individuals and therefore measure should be taken to ensure that data is not presented incorrectly. I hope that I have got them around the right way.
http://www.experiment-resources.com/type-I-error.html
http://www.lifesci.sussex.ac.uk/home/Zoltan_Dienes/inference/Neyman%20Pearson.html
It is important for psychologist to protect their participants when conducting research and to take their wellbeing into consideration. Ethic’s was not a big issue prior the famous research conducted Milligram on obedience. Since the negative reviews on his study ethnic guide lines were brought in to prevent further criticism. The British Psychological Society formed a code in conduct in which researchers were to follow when conducting a research, the guidelines included; Informed consent, Debrief, Protection of others, Deception, Confidentiality withdrawal from the research. *
Informed consent involves telling the participants the research is about and the asking for their permission to continue with the study. However there are difficulties when getting informed consent as it is impossible to tell the participants what the study is without them changing what they otherwise would have done; therefore they can explain and give an overview. Participants have to be issued with a number of contacts if there are any issues following the study and the procedure that have to take part in. For example within quantitative research it is hard to obtain consent from the individuals being observed as many observe in a wide area watching the behaviour in everyday life.
Debriefing is also an important procedure and one that has to be followed. Debrief is the procedure where the researcher has to provide the participants with an overview of what the aim of the study was and the purpose in doing so. Participants have the right to ask questions and the researcher must answer them truly.
Protection to participants is another guideline that has to be followed regarding ethics. Within this area researchers have to ensure that participants will not experience any distressing occurrences within the study and after it.
Deception is questioned on a occurring bases within the psychological testing field. If for example researchers do not tell the participants what they are testing then this is known as deception. However if the participants knew exactly what was being measure then their behaviour might change and the choices they make within the study might be influenced. For example if participants in the Obedience study knew that the other person was not actually being electrocuted then it might have changed the actions some of the participants made, more than 65% might have continued or less might have ignored the experimenters voice of authority. Therefore it is hard to establish what participants can be told and what they cannot. Deception is present in many research reports.
Protection to the participant is also essential and researchers ensure this is done by protecting the individual’s identity and coding their part in the experiment. They must ensure that is confidential at all times and the data is locked away.
Finally participants have to be given the opportunity to within at any given time without any penalties or being asked to stay.
Hope that everyone has a good Christmas break!
Happy Christmas! J
(*) http://www.simplypsychology.org/Ethics.html
(**)http://www.wadsworth.com/psychology_d/templates/student_resources/0155060678_rathus/ps/ps01.html
Hey Julie here are the commets for week 10 and 11
http://leilla92.wordpress.com/2011/12/09/is-it-possible-to-prove-a-research-hypothesis/#comment-50
http://psuc98.wordpress.com/2011/12/09/is-it-possible/#comment-24
Hope that you have a nice christmas and new year
hope we have you as a T.A next semister. ]
Happy Chirstmas
Jodie! 🙂
statsjamps 