Personal tools
You are here: Home → Core Programme → Archived material → 2010-2011 → Data collection and analysis 2011
Log in


Forgot your password?
New user?
 

Data collection and analysis hotseat

A hotseat conversation with Dr. Ian Terrell

Imagine you had to collect data on the different types of cacti...Cacti Four

A hotseat by Dr. Ian Terrell


Part 1: Data collection 


Your opportunity to discuss your thoughts and concerns on the issue of data collection. Before you start you may like to consult the material present here.

 

Aim of the hotseat:

This hotseat will be shaped by your contributions and queries. Nonetheless, the aim is to familiarise ourselves with the criteria and aim of data collection.

As a basic truth you will need to show that you can systematically collect data using appropraite techniques that are likely to lead to accurate and truthful portrayals or reality. You will also need to demonstrate that you have been selective, and self critical.


Part 2: Data analysis


Data analysis; all i need is a calculator...TRUE or False?

Preview from the data analysis research session:

Watch the rest of the videos here. 

Your aim and the MA Criteria:

 

Show that data collection and analysis has been systematic and structured.

Show that you have evaluated the methods you have used and are aware of their strengths and weaknesses.

Research is not opinion, belief, or faith in what is true but the systematic collection and analysis of data. So how do you explain the procedure you went through to justify calling it "research" and what issues that may lead to doubts in your accuracy, arise?

A more detailed account of the above criteria and extra resources can be found here.


question time Please contribute your thoughts, questions and innovations.

Document Actions

Ask a question: Raise and issue

Posted by Ian at Jun 01, 2011 11:17 AM
We have spent a lot of time talking about data collection this year.

You need to show that you have research skills. This means that you are systematic, You are self critical. You are aware of potential bias that creeps in. You guard against bias as much as possible. Your data collection is transparent. We can see the steps (procedure).

What problems arise for you?

Ask a question: Raise and issue

Posted by Christiana Kefala at Jun 01, 2011 07:41 PM
Hi Ian. My main data collection tool is a semi-structured questionnaire. The majority of the questions i have used are closed-ended (rating in order of preference) and these i believe i can analyze as quantitative data. However, there are a few questions that are open ended. I used these because i am researching students' opinion on praise and needed to get more details on their beliefs and attitudes. I am not sure how i can analyze the results i obtain from these questions. Summarize the main beliefs and attitudes that come up? Categorize the basic facts? Could you suggest any reading materials on this subject? Thank you

Ask a question: Raise and issue

Posted by Ian at Jun 03, 2011 01:00 PM
Sounds like good data collection Christiana mixing qualitative with quantittative.

You say your questions are about their beliefs so I guess the analysis and account has to be on their beliefs. They may have categoreries or you can group the beliefs under different headings.

Material on alaysing data is here http://midwhebonline.org.uk/[…]/analysing-data

All the main texts on data collection mention analysis see http://midwhebonline.org.uk/[…]/reading-list

Raise and issue

Posted by Jessica Jacob at Jun 22, 2011 10:22 PM
Hello Ian,

I am currently looking at the effect of lighting and visual tracking. I was wanting to work of a small group of kids in different lighting environments to see how and if the lighting effects them in any way.
I wanted to link this up with visual tracing issues. Do you have any ideas?

Jes

Research journals and Reflective Journals

Posted by Ian at Jun 01, 2011 11:29 AM
Many people use these. I don't see enough of them in appendices and I do not see enough explanation about the procedure in the Methodology and Procedure section (or plan section for DPP modules)


Have you seen http://www.audiencedialogue.net/journal.html

Or http://www.teachingexpertis[…]ective-learning-journal-642

Or
http://www.youtube.com/watch?v=e-KyI4jtvWc

Are you using reflective journal as a practitioner researcher?

Has anyone found?

Posted by Ian at Jun 01, 2011 11:31 AM
Has anyone found any good resources on "critical incident analysis" procedures?

Has anyone found?

Posted by Chloe Doherty at Jun 11, 2011 05:20 PM
http://www.bioscience.heacademy.ac.uk/[…]/beej-14-4.pdf

someone using it - good?

Has anyone found?

Posted by Jessica Jacob at Jun 22, 2011 10:34 PM
I found this sit helpful with clear simple explanations. http://www.usabilitynet.org/tools/criticalincidents.htm


Has anyone found?

Posted by Anthony Skip Basiel at Jun 23, 2011 07:49 AM
Hi
I have done some work with CIT.
Please have a look at my thesis in eLearning Design at
http://abasiel.wordpress.com
or contact me directly at abasiel@gmail.com
As I'm one of the tutors I'm happy to set up a phone / web video conference discussion too.
yours
Dr Anthony 'Skip' Basiel

representing data

Posted by Jacqueline Schembri at Jun 02, 2011 10:04 AM
My question is very similar to Christiana's.
In a case study of four subjects, what is the best way to represent data from open-ended questions in interviews ?
In the analysis of data I have written out all the responses under each question and given them codes. But I am a bit stuck with displaying this information. Any ideas ?
Thanks

representing data

Posted by Ian at Jun 03, 2011 01:03 PM
Cristiana could draw a map of beleifs and values. The landscape could be the mainstream, the way off beam in the distance, the alternatiev view. You misght draw a diagram of forces or tensions. You can be inventive.

Of course there is the structured narraitive of explanation.

In your case you say you have four cases- Can you write about each separately and then write an overview/comparison. Yopu would needs this second level analysis.

Critical Incident Analysis

Posted by Ian at Jun 03, 2011 01:20 PM
Just added a fantastic virtual tour of 3 guides to Critical Incident Analysis.

They are at http://www.midwhebonline.org.uk/[…]/collecting-data


Why dont you recommend some materials on research, data collection and analysis?

material on research

Posted by Christiana Kefala at Jun 03, 2011 08:09 PM
Ian, thank you for the recommendations. I think i will use basic headings to categorize the main ideas so that they can also be related to the quantitative data and then possibly add a few of the answers that really stood out and represent the sample. I searched online for other materials and found a good portal for all kinds of data analysis. So here it is, hope it helps:

http://lrs.ed.uiuc.edu/[…]/tselink.htm

weighing the evidence

Posted by Jacqueline Schembri at Jun 04, 2011 02:52 PM
Hi Ian,

Can you explain what is exactly needed, i.e what you expect to see in the assignment in the section 'assessing the weight of the evidence'. I have seen the example weight of evidence (for impact of CPD) but can't quite understand.

Thankyou

weighing the evidence

Posted by Ian at Jun 06, 2011 08:43 AM
Your analayis of the data can contain many things.

One section might be foused on assessing the weight or strength of the data that you have to support the main findings. Some findings may have strong data. Others may have weaker data.

You would show that you have been self evaluative of your data.


Weak data might be:
Not a lot
Conflicting, wide ranging or lacking in real focus
Perceptions of rather than observations of
Stated Beleifs that might have been infleunced by the circumtsnace or procedures

weighing the evidence

Posted by Jacqueline Schembri at Jun 07, 2011 03:48 PM
Thankyou Ian for your explanation.

selecting a sample size for collecting of Data

Posted by Lorrise Denham-Walters at Jun 10, 2011 11:10 AM
Hi all
Just concerned about my sample size. My topic has to do with sensory integration and the autistic pupil. There are about 32 pupils diagnosed with autism of a total population of 120. I only currently have direct access in lesson time to 5 pupils and 3 (year 11) will leave in two weeks time. What is a suitable sample size to use to make my data valid and reliable.
If I have to use other staff members to observe pupil behaviour, can this affect the validity of my research.
Regards
Lorrise

selecting a sample size for collecting of Data

Posted by Ian at Jun 11, 2011 08:33 AM
For any kind of statistical significance work a sample of 30 plus is normally required.

Of course it depends on the total population size and what you are generalising to. you may wish to generalise to 340 million school children! and not just 120 that happen to be at you school. This would need more in the sample. Your eight a a sample of all autistic children. Non of this is valid and reliable sampling if quantittaive work is carried out.

Smaller samples are acceptable. In your circumatmces the majority of your work should focus on in depth detail derived from qualitative work. With 5-8 you can even write up "case records" of qualitative interviews and then write a chapter drawing together common themes.

 

Ecological validity

Posted by Chloe Doherty at Jun 11, 2011 05:10 PM
I thought this had some good things to say about Ecological Validity

http://etd.library.vanderbilt.edu/[…]/Voorhees.Thesis.Final.pdf

Sensitivity and Bias

Posted by Christina at Jun 12, 2011 10:38 AM
Having read the many useful comments I remain slightly confused as to how to remove bias and personal allegiences when asking colleagies to comment on the input and value of classroom assistants to their lessons through a questionnaire ,some may feel a negative comment may question their direction of their assistant ,some may feel that by promoting the value of LSAs in their lessons that they are undermining their own value .Any ideas ?

Sensitivity and Bias

Posted by Ian at Jun 17, 2011 11:06 AM
Remove may be too strong a term when admitting to them explicitly would be better. You are after all committed, professional, an improver etc. Thats why we explore bias, beliefs and values, in professional autobiographies.

Limiting bias is acheived through systematic, transparent, and skillful data colelction and analysis. Sytematic menas both step by step and supported by the research literture as what othesr do. This means planning questions etc. as well as a spiirt of enquiry, self criticism, self evaluation and an opneness to other possibilities. Balance has to be shown. Descriptive data needs to be reported first before judgments about that data and explanation are offered. The latter should contain alternatives.

The practitioner researcher needs to emphasise the need to know to improve contribution

Sensitivity and Bias

Posted by Ian at Jun 17, 2011 04:00 PM
Thats why the step by step process of data colelction needs to be shown so we can see the bias that crept in at that stage.

Analysing data

Posted by Ian at Jun 17, 2011 03:45 PM
Your procedures for data colelction need to be transparent and systematric but so does the anlaysi of thsat data.

Dont forget to look at the materials here
http://midwhebonline.org.uk/[…]/analysing-data

Analysing data

Posted by Christina at Jun 18, 2011 12:13 PM
Very useful materials ..many thanks.

Wikipedia

Posted by Ian at Jun 17, 2011 03:58 PM
I am having a campaign supporting wikipedia against the luddites that complain its not academic (believeing that every other text is 100% corrrect!) so, look at http://en.wikipedia.org/wiki/Data_collection

Tell me what you see that you disagree with.

Collection and Analysis for Practitioners

Posted by Ian at Jun 22, 2011 09:28 AM
Being Practitioner Researchers affects data colection and analysis?

What are your thoughts on this?

I am thinking: Data Colelction: Emphasis on systemtic journal writing of everyday events. More emphasis on description and then analysis of events as live processes. The need for subsequent second level analysis. Critical incident analysis.


Analysis: Much more emphaisis on experience building concpetual models to structure analysis. Thank heavens we told you about profesional autobiographies, beleifs and values and subjective accounts!

Collection and Analysis for Practitioners

Posted by Ian at Jun 22, 2011 09:30 AM
Do have a look at http://midwhebonline.org.uk/[…]/collecting-data


Especially the precise way of "Critical Incident Analysis"

Reflection as an analytical tool

Posted by Ian at Jun 22, 2011 09:56 AM
Reflection is much banded around and sometimes seems to mean "just thinkin g about". I think it is more than that.

My comment above mentioned a second phase of anlaysis of incidents. I am reminded of the work of Argyris shown by INFED at (http://www.infed.org/thinkers/argyris.htm)
also Schon (http://www.infed.org/thinkers/et-schon.htm)

and the fantastic wikipedia
http://en.wikipedia.org/wiki/Double_loop_learning referencing Bart simpson.


Does the concpet of "double loop" explain the 'analysis" process?

Validity

Posted by Ian at Jun 22, 2011 03:04 PM

Remember that your research and its trustworthiness, its truthfulness can be measured by its "validity". Data colelction procedures can be valid becuase they are referneced to existing researchers in the procedures section. The description of the data can be "valid" ie the data does indeed look like the account.

Explanation is another matter. Explanatory validity is more tentative and there may lotts of different possible explanations.(Ultimately it may be the enthusiasm of the teacher that led to the impact)

Face validity is alos used as a measure- it sounds as though it makes sense.

However, in a community of practitioners, where they agree what to beleive is the case, what steps in data collection and analysis should be taken to ensure that their is genuine research discovering new insights?

Practitioners are in danger of colelcting their own data, using their own categories to come up with the way they already see the world. Where is the criticality coming from in the analysis?




Validity

Posted by Jessica Jacob at Jun 24, 2011 01:20 AM
Hello Ian,

I am currently looking at the effect of lighting and visual tracking. I was wanting to work of a small group of kids in different lighting environments to see how and if the lighting effects them in any way.
I wanted to link this up with visual tracing issues. Do you have any ideas?

Jes

Validity

Posted by Ian at Jun 25, 2011 01:48 PM
Sounds like a small scale observational study linked to informal semi structured interviews?

It would be nice to see some systematic observation records as a data set.

Diagram

Posted by Ian at Jun 25, 2011 01:49 PM
Is it possible to show your data collection and analysis as a seriesd of steps in a diagram?


Why do so few people do this?

Value of the web video clips

Posted by Anthony Skip Basiel at Jun 29, 2011 02:17 PM
Hello everyone -
As part of our continued evaluation - please do let us know your thoughts about the web video clips.
You can post them here and email Tarek with any thoughts you don't want public on the conversation area.
yours
Dr Anthony 'Skip' Basiel

Learning log

Posted by Scott Keough at Jul 02, 2011 01:30 PM
Hello everyone,
Just a quick question relating to the use (or not) of a learning log/journal.
Being as my study evaluates the effect of the implementation of a curriculum which I have been involved in developing, I decided not to use my learning log in my data analysis in the end. I have justified this is my methodology section as an attempt to mainly collate data produced by others (teachers and students) to try to limit the possible effect of bias and to strengthen the descriptive validity of my work.
Instead, I reviewed curriculum maps and assessment procedures, interviewed three members of staff, four students, collated 60 questionnaires and analysed attainment data so that I could triangulate between those data sets to try to provide a less bias, more neutral evaluation of the curriculum.
What are you thoughts?

Learning log

Posted by Ian at Jul 02, 2011 05:48 PM
Pity

A good well wqritten journal, well sytematically analysed would be good to read and enhance the work especially if it showed analysis and conceptualising.

Learning log

Posted by Scott Keough at Jul 02, 2011 06:16 PM
I am well over on the word count as it is and this seemed to be the most disposable part especially as I felt that excluding it may enhance validity. I was concerned that I may have too much data

Word Count

Posted by Ofori-Kyerah at Jul 03, 2011 11:45 AM
I am concerned about the word count. It appears the information I've put down so far is relevant to the study findings, however, it far exceeds the limit no matter how hard I try to cut down. Will I be penalised for exceeding the word limit?

Barbara

Word Count

Posted by Ian at Jul 05, 2011 03:53 PM
Yes- through sheer boredom no doubt.


You only have to pas the MA by shoing the criteria. Writing more does not shwo the crieteria but extravagant over indulgence. You should edit.

Learning log

Posted by Ian at Jul 05, 2011 03:54 PM
The log could be put in an appendix and would not count as words. It then needs to be referenced. A good log would show the criteria well.

Learning log

Posted by Ian at Jul 05, 2011 03:55 PM
In the end the views of your supervsiuor and first marker are the imporatant ones. Check it out with them and see if your work makes sense.

Learning log

Posted by Scott Keough at Jul 07, 2011 01:30 PM
I have spoken to my supervisor who has said that I should justify not using my log using references but to put a couple of pages as an appendix to show it was carried out.
Thanks
Scott

Data Collection & Analysis

Posted by Zeljko Kovacevic at Jul 28, 2011 12:22 PM
I have tried to be systematic and structured in my research, but inevitably there was a time when I felt that I am steering and guiding the participants towards the certain answers. Also, some students were asking me what to answer, despite the fact that I have told them that we expect their truthful answers and that they will not be judged by what they have answered. When I have started writing the questions, I already had in my mind the overall picture which I expected to be painted based on my own experience and informal talk with the students and teachers. What I could not have known are the exact percentages and correlation between data collected.
When asking questions and opinions my fellow teachers, I was more generic (open type of questions) than with the students (closed type of questions). Sample size for students was 240 and on the other hand only 6 for teachers.
As this was my first serious attempt of data collection and analysis, I am overall happy with results and I will be able to learn from my mistakes when I do my next research.
« June 2013 »
June
MoTuWeThFrSaSu
12
3456789
10111213141516
17181920212223
24252627282930