EER500

Interviewing

My current plan for my assignment is to have some initial questionnaires and then follow that up with some exemplary case studies.  I envisage that the case study information will be drawn from interviews and possibly student work samples.

Based on the information in Bryman (2015) I am likely to take a semi-structured interview approach.  I’m interested in the possibilities of giving the interviewees the questions in advance, so they can think about their answers. I think this would lead to more considered answers (like in an email interview) but also has the advantage of spontaneity.  I would probably limit these interview case studies to three, half-hour interviews (which then would take about 3 hours each to transcribe).  Ideally these would be face to face interviews, but depending on how I do my sampling, the participants may be too remote to access physically, and in that case an interview via Skype or similar, providing a face to face live online interview.

Combining Qualitative and Quantitative Research

The method I intend to use is called a Explanatory Sequential Design where I use quantitative data to explain the qualitative data I have gathered, both the quantitative data is weighted as more important, and comes first.  The quantitative data will also be used as a method of identifying potential case study participants.

References

Bryman, A. (2015). Social research methods. Retrieved from https://ebookcentral-proquest-com.ezproxy.csu.edu.au

Using Existing Data

Part of (what I hope will be) my research project proposal will draw from existing data – NAPLAN Data. So I chose this chapter (Bryman, 2015, Chapter 14) as my starting point (in two weeks of readings where there are seventeen readings to choose from).

Bryman (2015) discusses secondary analysis of data sets and details how to access that in the UK, but I wonder if we have anything comparable in Australia.  I know we have NAPLAN data and Census data, but I wonder if there is anything else publicly available and/or accessible?

The benefits of using existing data are that a small scale research project, such as a student project, will have access to a larger data set than they could survey in the time available, official statistics tend to be an unobtrusive method of researching, and the large sample sizes mean that the samples are more likely to be representative.  Some of the downsides are the manipulation of public data to make departments (such as the police department) look better, and how representative that data actually is, such as not all crimes being reported.

Using NAPLAN data falls somewhere between secondary analysis and official data because students sitting NAPLAN do know they are being tested, but it is official Government data.  My assignment will also use ABS census data for sampling purposes and this definitely comes under official statistics, although, again, people know they are being “examined”.

Self-Administered Questionnaires

I intend for the first part of my research project to include a self-administered questionnaire.  I would post this out, including a stamped, self-addressed envelope, but also include a link to an online version of the survey to mitigate the problems with postal surveys being lost.  The benefit of a self-administered questionnaire is that there is no researcher bias (either in influencing the answers given or the interpretation of the answers), the answers are returned relatively quickly, are easy to code afterwards and take much less time than face to face or telephone interviews.

References

Bryman, A. (2015). Social research methods. Retrieved from https://ebookcentral-proquest-com.ezproxy.csu.edu.au

I found this example of what happens when correlation and causation are confused.

The chapter of the textbook (Ch 4) (Bryman, 2015) is more relevant to the second assignment, I feel, in formulating a research question and establishing how you will examine it.  I hope to review this chapter again in preparation for the second assignment.

I somehow got 12/13 on a ten question quiz for this chapter.  I got 9 out of 10 questions correct.

According to Punch (2005) my question development process has been inductive – I have started from a very specific question and then moved more generally.  So, for me, I want to look at whether the presence of a trained teacher librarian impacts NAPLAN scores for the school.  This has lead me to look at research in Australia and then globally about the impact of a librarian in a school.  At this point I am envisaging that my research question will become a little more specific, to a specific geographical area or a socio-economic bracket.  I am thinking I will see if the presence of a teacher librarian is still beneficial to schools that are in a low-socioeconomic area.

References

Bryman, A. (2015). Social research methods. Retrieved from https://ebookcentral-proquest-com.ezproxy.csu.edu.au

A fascinating and relevant case has just come to light in the last few days.  Scientists storing their research data on the cloud-based-service Dropbox had their research data used, without their consent, for another research project. This highlights one of the four main areas of ethical concern with regards to research – informed consent. Other ethical issues involve avoiding potential harm to participants (such as the Emotional Contagion study conducted by Facebook), maintaining privacy of participants and whether deception is involved (Bryman, 2015).

I recall a story from my undergraduate days, where ethnographers were studying a population group and had to remove themselves from the situation because the nature of the questions they were asking were impacting the beliefs of the group.  The study uncovered a practice where parents left their infant children alone all day, buried in sand in a large bag.  I have been unable to find the original research paper but I did find reporting on it (Branson, 1990).

I got 9/10 on my chapter quiz.

Update

Our readings (Bryman, 2015) indicated that ethical issues can include both large things (like lack of consent or confidentiality) and small things (like under-quoting on the length of time a survey will take to increase participation).  So when I received a survey in my inbox today, I thought I would test it out.

The first page of the survey upgraded this to “10-12 minutes”, which I thought was interesting, however, in the end, I don’t feel this was misleading as the survey took me seven minutes to complete.

References

Branson, L. (1990, March 25). Peking tackles sandbag way of babycare; China. Sunday Times[London, England]. Retrieved from http://link.galegroup.com.ezproxy.csu.edu.au/apps/doc/A117290935/EAIM?u=csu_au&sid=EAIM&xid=361d0f63

Bryman, A. (2015). Social research methods. Retrieved from https://ebookcentral-proquest-com.ezproxy.csu.edu.au

Quantitative Sampling

I had an idea in mind for how I was hoping to form my research proposal, and (of course) had an idea for the topic I was intending to focus on.  However, I have come to realise that sampling is far more complicated than I had hoped, especially if you are after probability sampling.  However, my plans were limited to what I thought was achievable for myself, rather than what I could theoretically propose, because I don’t actually have to carry out this research project! So even if I personally could only feasibly conduct research within my local area, I am not constrained by that for my assignment.

This leads to how I would define the criteria for inclusion rather than “schools in my local area”.  Should I balance for socioeconomic area, so that all “walks of life” are represented?  Should I look at schools with a variety of levels of academic achievement?  Should I focus on one area – such as schools from a low socio-economic background, to justify that the inclusion of a teacher librarian in a school in a less affluent area, as a cost-effective measure?  So many possibilities?

What factors would you take into account in deciding how large your sample should be when devising a probability sample?

Time and cost are the two key factors when deciding how large your sample size should be, although you must also bear in mind that the larger the sample size, the closer it will be to being a truly representative sample.

In what circumstances might you employ snowball sampling?

Snowball sampling is more often used in qualitative research, but is particularly useful in hard to access populations, such as research involving people who engage in illegal acts, or in stigmatised populations, for example, a survey of drug users or of sex workers.

I got 8/10 on this chapter test (Ch 8).

Qualitative Sampling

I did not cover this subject in detail, but only skimmed, as it is not the intended focus for my research project.  However, I learnt that the focus for qualitative research is purposive sampling, rather than probability sampling and that other things can be sampled other than people, such as organisations or documents.

References

Bryman, A. (2015). Social research methods. Retrieved from https://ebookcentral-proquest-com.ezproxy.csu.edu.au

After reading the textbook chapter on Quantative research (Bryman, 2015), I have a greater understanding of the complexity of such research, which is important since I intend to have quantitative research as the basis of my research proposal at the end of this subject.  Based on the decision tree in the readings (McMillan & Wergin, 2010, p 5) I am guessing that the style of research I hope to propose is Ex post facto design, or possibly Comparative – it’s hard for me to work that out at this stage. There is also a useful set of questions for evaluating quantitative research (McMillan & Schumacher, 2006, p 32) which I intend to use to examine research I read but also use to guide my research proposal. I’m pleased to have gotten 9/10 in this chapter quiz.

References

Bryman, A. (2015). Social research methods. Retrieved from https://ebookcentral-proquest-com.ezproxy.csu.edu.au

This post is simply serving as a place for me to collect my thoughts about my current idea for my literature review and research proposal for EER500.

I am still currently working on the research question of whether the presence of a trained teacher-librarian has an impact on NAPLAN Literacy scores, and whether or not this is a more cost-effective method of raising average NAPLAN scores than a teacher employed for remedial reading instruction.

My starting points for my literature review will be:

Parliament of  Australia, House of Representatives committees. (2011) Inquiry into school libraries and teacher librarians in Australian schools.

Lupton, M (2016) Adding value: Principal’s perceptions of the role of the teacher librarian,School Libraries Worldwide, 22/1

Hughes, H. (2013) School libraries, teacher librarians and their contribution to student literacy development in Gold Coast schools. Research report.

These papers should then connect me with other articles, both in seeing which papers they cited and in examining who has since cited these papers.

Research Methods

I think quantitative research is the best method suited to this specific question, although I think I would generally lean towards the perspective promoted by qualitative research.  I would use a combination of surveys and publicly accessible NAPLAN data to research this question.

Challenges

Primary challenges with this research actually being practical include:

• the busyness of school executive leading to a potentially poor return rate.
• the lack of specificity in the publicly accessible NAPLAN results.
• value for money should not be the only factor in determining a course of action.
• the benefits a trained teacher-librarian brings to a school cannot be limited to the potential for improved NAPLAN results.

In the course of these readings (I wish I could remember which one!) I came to realise that the history of peer-review as the gold standard of publishing research articles is only roughly 100 years old, which was surprising to me.

Qualitative Research engages with people – usually interview subjects – more deeply, to provide a narrative view, and it typically includes a statement acknowledging the backgrounds and biases of the authors (Creswell, 2012). It generally relies less heavily on a literature review, which may only form the introductory portion of the article (Creswell, 2012).

Qualitative research differs greatly from quantitative research in many aspects, one of them being the replicability of the study, as qualitative research is, by nature, more subjective than the objective quantitative research.  Hence, there are many different theories as to how to best judge the quality of qualitative research (Bryman, 2015).

Qualitative research doesn’t easily lend itself to a clear definition or a distinct, predictable process and so has lead to criticisms of lack of transparency, bias and lack of general applicability (Bryman, 2015).

Can qualitative testing be employed in relation to hypothesis testing?

While qualitative research is generally conducted to produce theories or concepts, it can be used to confirm (or contradict) an existing hypothesis (Bryman, 2015).

I received 8/10 on the chapter quiz.

References

Bryman, A. (2015). Social research methods. Retrieved from https://ebookcentral-proquest-com.ezproxy.csu.edu.au