Homework for my TA

Dear Wendy,

I hope you had a nice Easter break, please find below the comments I have made this week:

1.http://sinaealice.wordpress.com/2012/03/25/research-into-cyber-bullying/#comments (sorry there is a word doc paste error! Could not find a way to edit it. Hope I can )

2. http://laf1993.wordpress.com/2012/03/25/do-you-always-have-to-read-the-instructions/#comment-44

3.http://lefeza.wordpress.com/2012/02/19/fasting-or-eating/#comment-30

4.http://zoe3112.wordpress.com/2012/03/25/bad-science/#comment-20

Kind Regards,

dsm1lp

Is it ethically ok to use internet sources as data for qualitative studies?

The web is full of many different internet sources; you can find written information everywhere, on websites, chat rooms, blogs and news feeds. However there are many ethical issues involved in using online data – such as consent.

The question is: Just because the information’s online and we have access to it (in most cases). Does it mean it is ethical to use this information? with or without consent?

Or

Does publishing this information mean users are already giving consent? Making it ethical?

The following web link looks at a question regarding the amount of sites built each year – http://css-tricks.com/new-poll-sites-per-year/(Coyier,2010). Threads on this site discuss this matter (logging in needed) – written by online users, some anonymous like “Me Me ME” (Coyier,2010), some not. If I wanted to look at qualitative information regarding the job titles of the commenters, would it be ok?

This leads me to suggest that it is not ethical without the commenter’s permission, based on the fact that this information is of a personal nature – It’s their views and details. Others may support the ethical use of this information because it’s openly published for everyone to access. However, reading what people have said is different to using what they have said. Therefore ethically, consent should be requested before information is used.

http://www.ljmu.ac.uk/administration/administration_docs/DPA_The_Provision_of_Consent_to_Process_Personal_Data.pdf

Getting this consent can be hard as the person discussing the topic may wish to hide their name or not give consent when asked. Importantly, the article above states that consent is not given by lack of requested response. In addition, the commenter may or may not be able to give it and therefore consent maybe needed by the parents. However, their parents may not know about the information they are sharing online and the child may be inclined speak to their parents regarding the matter.

http://www.teachingexpertise.com/e-bulletins/legal-consent-when-and-how-it-needed-school-3749

Recently, watching a tv show (The Anti-Social Network – http://www.bbc.co.uk/programmes/b01dwg1n) regarding bullying online, made me consider whether it would be ethical to use chat room data. Would signing into a chat room which discusses online bullying and using the data be ethical?

One chat room I found online was http://www.chathour.com/chatroom/ant-bullying. This chat room required you to log in, supporting the idea that it would be un-ethical to use this chat room information. Based on the fact that people are logging in to chat about bullying, not logging in to a chat room designed to collect research on the matter. The login factor makes the information more private and protected – http://wiki1.dovecot.org/LoginProcess .Suggesting that the users may want to chat with other users who may relate to their experiences, rather than just anyone. Therefore using their data could be considered as a privacy issue – if they are unaware. (Secton 3 – http://www.privacyrights.org/ar/Privacy-IssuesList.htm#onlineprivacy)

Evidence shows that in reality Information online is (mostly) accessible; however this information should not be used in research without the users being asked. Information such as comments which are open online could be argued as ethical to use because the person knows the information is online. If the site needs you to login, then the user of the information may wish for this information not to be directly seen by everyone – suggesting that gaining consent maybe even more ethically important. Overall, consent should be asked for when it comes to the use of chat room information and blog comments .Using sources online without consent, specifically when chartroom login is need, is therefore suggested unethical.

References

Coyier, C. (2010). New Poll: How many websites do you launch per year? Retreived from: http://css-tricks.com/new-poll-sites-per-year/

LJMU Data Protection Office: Provision of Consent to Process Personal Data and Data Protection Statements Retrieved from: http://www.ljmu.ac.uk/administration/administration_docs/DPA_The_Provision_of_Consent_to_Process_Personal_Data.pdf

Mills, K. (2008). Legal consent: when and how is it needed at school? Retrieved from: http://www.teachingexpertise.com/e-bulletins/legal-consent-when-and-how-it-needed-school-3749

Richard Bacon (2012) BBC THREE: The Anti-Social Network, Retrieved from: http://www.bbc.co.uk/programmes/b01dwg1n

(2012). Chat Hour, Retrieved from: http://www.chathour.com/chatroom/ant-bullying

Sirainen, T. (2010) Login Processes, Retrieved from: http://wiki1.dovecot.org/LoginProcess

Privacy Today: A Review of Current Issues, Retreived from: http://www.privacyrights.org/ar/Privacy-IssuesList.htm#onlineprivacy (Revised March 2012)

Homework for my TA

Dear Wendy

I have commented on the following blogs this week,

1.http://kdjhg.wordpress.com/2012/03/07/using-animals-in-psychological-research/#comment-60

2.http://frasersstatsblog.wordpress.com/2012/02/05/can-alcohol-improve-your-revision/#comment-68

3.http://prpij.wordpress.com/2012/03/09/an-interesting-concept-from-a-popps-prepared-speech-should-have-clive-wearing-been-tested-on-if-he-couldnt-remember-giving-consent/#comments

4.http://bellaistoocool.wordpress.com/2011/10/13/is-it-right-to-use-children-is-studies/#comment-52

Kind Regards

Why is the “file drawer problem” a problem?

File drawer problem? What is this and why is it a problem…

Publication bias is notion related to the idea of the file drawer problem where research, sometimes unpublished, can influence people thoughts causing bias views. The file drawer problem is the idea that results are only really discussed/published when they are significance (under .05), suggesting that this results in people being bias toward significant ideas. The file drawer problem influences people – negatively because it makes them inflexible with their view, leading to an inability to see the other side due to the lack of information presented.

The file drawer problem must also have implications in the real world, not only in research. Let’s think about the information we read. We tend to read and share information about; holiday destinations, schools for our children and make-up. This shows the benefits of us going there, sending our children to that school or wearing that make-up. We generally, do not find easily available negative information, suggesting bad things regarding these matters. This causes bias in us thinking about the good things.

Ok, so now we know what the file drawer problem is, let’s consider what research suggests about the drawbacks of this idea.

Research by Scholey and Harrison (2003)* suggests that the accessibility of research and findings plays a role in the effectiveness of publication bias influencing us. If only a few articles are widely available, which lead to one viewpoint and others are not accessible on the other viewpoint. Then this can restrict the possibility of us to getting a clear idea of the research, which can result in people making flawed evaluations due to using small or large sources of one side information.

http://www.nature.com/bdj/journal/v194/n5/pdf/4809923a.pdf*

This is also show within the real world. For instance, your friend tells you in an email about a good school her son goes to and you read this schools report online. You are convinced and think that this is the best school for your son. In fact you have not considered the information your friend and the report has not mentioned about this particular school, due to it not being given directly. Looking at the file drawer problem in the real world, we can see that this bias can cause major errors in our decision and choices.

From looking at a letter published in Schizophrenia Research 90 (2007), it suggested that the file drawer problem is a problem within meta-analysis. As it cause problems with generalizability, due to the issue of hard to find research. To address the issue and make sure all research is found, this letter discusses the uses of computer when it comes to searching for articles. However, the vagueness of abstracts is suggested to cause problems still.

http://0-pdn.sciencedirect.com.unicat.bangor.ac.uk/science?_ob=MiamiImageURL&_cid=271176&_user=899436&_pii=S092099640600404X&_check=y&_origin=article&_zone=toolbar&_coverDate=28-Feb-2007&view=c&originContentFamily=serial&wchp=dGLbVlS-zSkWA&md5=7020f986205021fccb34ae29a1b0dd14/1-s2.0-S092099640600404X-main.pdf (letter in Schizophrenia Research 90, 2007)

From looking at the file drawer problem and its issues, we can see that a good amount of varied information needs to be obtainable to reach an unbiased perspective. If information is concealed or only subjective information is exhibited, this can have an impact on people’s verdicts of a given situation or idea. These influences lead to false impressions based on our narrow perspectives.

References and Further Information:

http://www.ma.utexas.edu/users/mks/statmistakes/filedrawer.html

http://www.nova.edu/~marker/2006/12/file-drawer-problem.html

Homework for my TA

Dear Wendy,

This week I have commented on the following blogs:

http://psucf0.wordpress.com/2012/02/19/a-case-of-bias-in-psychological-research/#comment-42

http://ppp1003.wordpress.com/2012/02/05/the-effects-of-music-on-the-brain/#comments

http://ppp1003.wordpress.com/2011/11/29/medication-or-meditation-the-best-way-to-concur-anxiety-disorders/ (Emailed Regarding)

http://statssam.wordpress.com/2012/02/05/types-of-reliability/#comments

(When I viewed this post it say’s ‘Your comment is awaiting moderation.’)

Kind Regards,

dsm1lp

What role does the null hypothesis REALLY play in the scientific process?

To understand the null hypothesis role, it is important to address exactly what the null hypothesis is and how the scientific process works with regards to this. The Null hypothesis is used within research as an alternative to our hypothesis – the main idea or the reasoning behind what we think has happened or is going to happen. However, the null hypothesis suggests our idea (hypothesis) is flawed, incorrect and is used to reject it. Furthermore, it can also suggest no effect or difference.

Even though, the hypothesis and null hypothesis can help researchers address their questions, it is important to remember that the null hypothesis does not give another explanation. It simply rejects or accepts the hypothesis. The null hypothesis is not directly the opposite of a hypothesis, due to the fact that it lacks an alternative explanation, to the idea that it has disregarded. Regardless of the fact that the null hypothesis works like this, it’s role is still vital. With further new hypothesis rejection, it is possible to cancel-out over time many explanations which are available for our results.

(Image retrieved from- http://www.null-hypothesis.co.uk/photos/2008_03_10_Null_Hypothesis.jpg)

Accepting the null hypothesis suggests to us to rethink our idea’s, which helps us take new steps towards other possible answers. However, rejecting our null hypothesis on the other hand can support our reasoning, which can be strengthened from replicated and new research.

The following diagram shows the steps within the scientific method. Within this method, the null hypothesis comes into play after we have done our experiment and are in the process of analysing our results. As you can see on the diagram the results analysis, can go two ways for our hypothesis. One of which, shows a suggested true result where the hypothesis is considered to a degree correct.  The other route, suggests that the hypothesis is untrue and an incorrect idea. With this false result the researcher could reject the hypothesis and accept the null hypothesis. When this happens the cycle starts again from stage 2, where the researcher needs to assemble a novel idea, which can be tested by its null hypothesis.

(Image retrieved from –http://s1.hubimg.com/u/4205792_f520.jpg)

The null hypothesis role within the scientific method is to show that the hypothesis is not an explanation for what the data is representing. Within the analysis we test the null hypothesis, to help us suggest whether our hypothesis could be a possible reason for the results we collected. The more we replicate the research and reject the null in testing – the more and more possible our initial hypothesis is suggested to be plausible.

Therefore, from looking into the null hypothesis and the scientific process, it is clear that the Null Hypothesis plays an important role in helping researchers to either; reject their hypothesis and go back through the stages or reject the null hypothesis which will support the hypothesis, leading to further studies to replicate the results.

References & Further Information

http://www.null-hypothesis.co.uk/science//item/what_is_a_null_hypothesis

http://www.sciencebuddies.org/science-fair-projects/project_scientific_method.shtml

http://www.windows2universe.org/people/scientific_process.html

Greer, B., & Mulhern, G. (2002). Making sense of data and Statistics in psychology. Hampshire: Palgrave.

Gravetter, J. F., & Wallnau, B. L. (Ed.8). (2009). Statistics for the Behavioural Sciences. Belmont: Wadsworth.

Homework for my TA

Dear Kat,

This week I have commented on the following blogs:

http://ksgs.wordpress.com/2011/12/09/deception/#comment-35

http://bartlettdan.wordpress.com/2011/12/09/must-the-need-for-documented-ethics-procedures-hinder-research-progress/#comment-19

http://em2904.wordpress.com/2011/12/08/do-ethics-prevent-possibly-helpful-research/#comment-32

http://ecpsychology.blogspot.com/2011/10/how-can-we-be-sure.html?showComment=1323466665372#c8293270086859858352

Kind Regards

 

Is it dishonest to remove outliers?

To answer this question there are many things which we need to consider. What is an outlier? Why should we consider removing or keeping it within the data? What would make it honest/dishonest to remove an outlier?

(Picture sourced -http://web.anglia.ac.uk/numbers/common_folder/graphics/fig6_single_box.jpg)

Outliers are seen within research either by looking at raw data or graphs. Researchers should think carefully about outliers, if they arise in research.

Research by Harmeling, Dornhege, Tax, Meinecke and Mu¨ller (2006) discusses how outliers can arise, suggesting that they can come from errors within the measuring of the data. Harmeling et al., (2006) also suggest that true outlier’s tend to have low probability and are likely, to be points which are out of sync with the other data points. The issue with outliers is the fact that they can have a negative effect on results, making the finding erroneous. However, it is suggested by Harmeling et al., (2006) that it is hard to decide what you should count as an outlier and what you should consider as a genuine result.

Furthermore, research by Hosseinalizadeha, Rivaz, Hedayatizadeh (2011) states the importance of detecting outliers within research, showing the effects they can have on finding. Their finding show, how the appearances of the graphs can change from results containing many outliers, to results without.

For a clear visual example please see page 344 on link:

http://0-www.sciencedirect.com.unicat.bangor.ac.uk/science?_ob=MiamiImageURL&_cid=270397&_user=899436&_pii=S1878029611001861&_check=y&_origin=search&_zone=rslt_list_item&_coverDate=2011-12-31&wchp=dGLzVlB-zSkWA&md5=cb83846dbcfb5380fd147f413925b916/1-s2.0-S1878029611001861-main.pdf

This research suggests that outliers can, in some cases such as this, be of interest if they are true scores (Hosseinalizadeha et al., 2011). However, if they are due to error, removal is necessary (Hosseinalizadeha et al., 2011).

Shaun Snapp (2010) discussed why outliers are removed (please see link below), he suggests that the removal of outliers can result in misleading information and false results. He also talks about the benefits to removing outlier and suggests that manipulating results can simply be done by removing outliers. (Link: http://www.scmfocus.com/demandplanning/2010/07/outlier-removal/)

From looking at how outliers come about and the effects they can have. Findings suggest that the removal of outliers is not dishonest, if it is done for the right reasons ie, the scores are erroneous. However, findings suggest that it is dishonest to remove them if they are true scores, as we would not want to be misleading. Therefore, if an outlier is going to be removed it must be justifiable, by the researcher as to why they have removed it.

( http://withfriendship.com/images/d/18083/Outlier-picture.gif )

References:

M. Hosseinalizadeh, F. Rivaz, & R. Hedayatizadeha. (2011). Importance of Outlier Detection in Spatial Analysis of Wind Erosion, Procedia Environmental Sciences 7, p.341–346 Retrieved from: http://0-www.sciencedirect.com.unicat.bangor.ac.uk/science?_ob=MiamiImageURL&_cid=270397&_user=899436&_pii=S1878029611001861&_check=y&_origin=search&_zone=rslt_list_item&_coverDate=2011-12-31&wchp=dGLzVlB-zSkWA&md5=cb83846dbcfb5380fd147f413925b916/1-s2.0-S1878029611001861-main.pdf

S. Harmeling., G. Dornhege.,D.Tax., F. Meinecke.& K.R. Mu¨ller. (2006). From outliers to prototypes: Ordering data, Neurocomputing 69, 1608–1618 Retrieved from: http://0-www.sciencedirect.com.unicat.bangor.ac.uk/science?_ob=MiamiImageURL&_cid=271597&_user=899436&_pii=S092523120500264X&_check=y&_origin=search&_zone=rslt_list_item&_coverDate=2006-08-31&wchp=dGLzVlt-zSkWb&md5=5aec7ae59de52351f08499e24e486ce3/1-s2.0-S092523120500264X-main.pdf

Other sources found:

http://www.itl.nist.gov/div898/handbook/prc/section1/prc16.htm

Should researchers conduct exploratory data analyses?

To address this question it is essential that we underline, what exactly exploratory data analysis is all about, by doing this it will allow us to develop our understanding of whether or not as researchers’ we should conduct research analysis in this manner.

Exploring data, can be the same as exploring anything else, not knowing what you are going to find/uncover until you start makes it interesting. When exploring data, it allows us to find out many things, for example simple things (some relevant, some not) like how many participants have been tested within the study etc. Furthermore, it allows us to find out complex things from data, for example how many students in the country gained a 2.1 in their last year at university compared to the amount that got a 2.2. All in all, exploratory data analysis is adaptive and it can use many methods of graphing to find out information and to see problems within the data. However, this analysis takes a different approach to others, in which the data shows us what it’s all about, rather than following assumptions to find answers.

Analysing data in this way can be good because it straightforwardly allows us to see what the raw data is telling us, using graphs and simple statistics, without making things too complex (which can cause small and major problems). In addition, exploratory analysis looks at graphs and summaries rather than statistical explanations, therefore it is suggested that this makes it a more in-depth approach.

However, quantitative data analysis on the other hand is different to exploratory data analysis, due to the fact that it uses tests to find out information, such as the significance of the results. Quantitative analysis is suggested to have a filter factor involved within it, which can cause problems. As assumptions are made, some of the data being focused on is in possible danger of being missed or devalued as it is reduced down. This shows a limitation within this quantitative approach that could be avoided through exploratory data analysis. An important fact, which we should not forget to mention, is the fact that both types of analysis could be used together.

Therefore, it should be up to the researcher to make a judgement on whether statistics are needed and whether they should use exploratory data analysis or not, depending on what the graphs/summaries suggest. However, exploratory data analysis seems to be a good method for looking at data as it can give a deeper understanding of the data and can give us a good clear visual idea of the results.

http://www.itl.nist.gov/div898/handbook/index.htm

Homework for my TA

Dear Kat,

This week I have commented on the following blogs:

http://francesdevine.wordpress.com/2011/10/14/is-it-wrong-to-remove-outliers-from-data/#comment-24

http://pumpkinqueenme.wordpress.com/2011/11/15/the-declaration-of-helsinki-and-the-five-basic-ethical-principles/#comment-40

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

http://em2904.wordpress.com/2011/10/14/why-is-reliability-important/#comment-22

Kind Regards,

dsm1lp