Tuesday, January 27, 2015

Shetland

Like no small percentage of the population, I too have been struck by a mild respiratory virus. It was sufficient to keep me in bed for one day, but I've been up and working (but not walking) since. Today I seem to have lost the power of speech.

While resting, I read the first four books of Ann Cleeve's "Shetland quartet". I'm not sure now how I found these; although a television series has been made of the novels, they haven't been broadcast here yet. It turns out that I read them in the wrong order, although this doesn't seem to have made much difference, apart from changes in the personal life of the books' protagonist, Jimmy Perez.

Viewed in order, the first book ('Raven black') takes place in Lerwick, the capital of the Shetland Islands; the second ("White nights") takes place in the far north of that island; the third ("Red bones") is set on a secondary island, Whalsay, and the fourth ("Blue lightning") is set on another island, Fair Isle. Thus one receives a geography lesson whilst solving mysteries.

Apart from the fourth book, I found the first three tedious reading. The plots are very slow, presumably reflecting the nature of the Shetlands, and there seem to be too many characters, most of which are related to each other. These books aren't police procedurals but rather who-dun-its, 'Blue lightning' especially. Unlike the Sophie Hannah books, these were at least readable, but I found them hard going and not interesting. I couldn't make any connection with any of the characters.

None of the books start with a murder: there's a certain amount of introductory material first. But all of them have a second murder about half way through; in fact, I found myself thinking whilst reading one of them that a second murder should occur soon - as it did, a few more pages in.

I'll give one of the books a second try in about a week: it may be that my general low feeling is affecting my reading and appreciation.

[SO: 3756; 3, 15, 36]

Wednesday, January 21, 2015

A pleasant surprise whilst writing 800 blogs

Following on from yesterday's blog, it became apparent that I would need the research questionnaire translated into Russian. I showed the Hebrew version to a few people near me and their faces took on the same expression that mine does when shown something when I'm not wearing my glasses.

There is no need to examine whether these people are even representative of the general Russian speaking population; if they can't read the questionnaire (or reading it would require an extortionate amount of time), then they are not going to answer the questions, which means that I lose people from my population.

So, as I suggested yesterday, I built a temporary version of the questionnaire by copying each sentence (questions, answers and surrounding text) into Google Translate, then copied the Russian text into a file, maintaining the same order. There are some answers which I didn't bother translating, as either they are solely numerical (e.g. 26-50%) or they are answers which repeat themselves (I agree, etc).

I am well aware of the fact that translations prepared in this manner are somewhat lacking (I've been on the other end of this, having had to correct a translation from Hebrew to English) and so I decided to ask more capable Russian speakers in my company to revise the Russian. I sent off four or five letters, and to my pleasant surprise, received affirmative answers within minutes. At least one said that the writer would be only too pleased to help. I suggested in the letter that the questionnaire be divided into sections so that no one has to translate/correct more than ten questions, but one person offered to correct the entire questionnaire.

It seems that I am not used to such kindness any more!



On another matter, this is now my 800th blog. With the help of the analysis program which I wrote a month ago, I can easily show what the hot topics of the last 100 blogs have been.

TagCount
DBA32
Holiday16
Health12
Personal12
Sorrento9
Programming8
Sicily7
ERP6
Cooking5
Films5
Israel5
Excel4
Kindle3
Mobile phone3
Statistics3
TV series3
19752
Computer2
Delphi2
Jewish holidays2
MIDI2
Maccabi Tel Aviv2
Meta-blogging2
Obituary2
Office automation2
SQL2
DCI Banks1
Enigma1
Firebird1
Food science1
Habonim1
In-basket1
Kibbutz1
Literature1
Olivia Williams1
Organisational behaviour1
Peter Robinson1
Psychology1
Richard Thompson1
Robert Silverberg1
Sandy Denny1
Swell Season1
Yoni Rechter1

No real surprises there. DBA and holiday together comprise 48% of the blogs (every 'holiday' blog was either also 'Sorrento' or 'Sicily', so these latter two labels don't count) and if I add in health, then 60% of the blogs are covered. So now we know what has occupied my attention since April 2014.

The interesting thing is that subjects which used to occupy me - programming, music, books and films - seem no longer to be the subject of attention. Does this mean that I am becoming more focused as I become older? Or could it mean that there are fewer and fewer pieces of new (to me) music about which I want to write?

Tuesday, January 20, 2015

Position paper submitted

A month ago, I wrote about writing a paper for the SEMS 15 conference which will be held in Florence in May. At some stage, I also wrote about how writing a paper has similarities with recording a song: the first two takes were ok, but not what I wanted. The third take was very good and this has been the basis of a slow but on-going dialogue between Felienne Hermans and myself over the past month. After a few overdubs (edits), Felienne advised me that it is time to submit the paper to the peer review process.

First, I had to format the paper according to the required standard. There is a document which gives the specifications, but it is also a 'live template' document. I'm not sure exactly what this means so it took me quite a while to get my paper formatted in the style required (Times New Roman 10, two columns, widths etc). In the end, I replaced part of the template's text with my text, which solved most of the problems. By hand, I formatted the title and the author details and pasted these in.

As the paper had originally been written for a different conference which has a different style (Arial 12), the new format shrunk the paper somewhat, enabling me to add a little information. SEMS also requires the references in a different format, so these had to be changed as well. So after no small amount of struggle, I had a version which contained what I wanted it to contain and formatted in the correct style.

I then had to contend with a website which allows one to update submissions to the conference. This took a little time but was fairly straight-forward. The site requires that the submission be in pdf format, so I had to create a pdf from the Word document which I had been editing (more overhead). Final gripe: the site consistently referred to me as "No'Am" Newman; I had to remove the apostrophe in order to achieve "Noam" - and now I have to hope that no one misreads this as 'gnome'.

I don't know how long the peer review process will take, but in the mean time, I will reserve both airplane tickets and hotel room (i.e. book but not pay).

I have also been working on the questionnaire. My supervisor does not like the idea of the Split Questionnaire Design, so I have been examining the questionnaire with a very critical eye to see which questions can be removed without influencing the results too much. The first step was analysing which questions belonged to which factors: it turns out that three of the first five questions were not connected to any factor and could be removed. They are 'good' questions but not necessary.

I then went over the different sections: there were a couple of questions in the 'Priority/Excel' section which didn't contribute anything and a few in the 'user satisfaction' section which were effectively duplicate. So after wielding a very sharp knife, I managed to reduce the entire questionnaire to 45 questions (and this is after adding a question about mother tongue). It shouldn't take more that 15 minutes for someone to complete the questionnaire, so it seems that the entire questionnaire will be used instead of the SQM (two 30 item questionnaires). This will be checked.

In a parallel track, people have been looking at the Hebrew version of the questionnaire and flagging parts which aren't sufficiently clear. Unfortunately, some people have taken a very long time (two weeks) to do so, which means that by the time I received their input, the wording of a question or answer had already been changed - or even dropped from the questionnaire.

I am also checking to see whether the average native Russian speaker who works with Priority can understand the Hebrew questionnaire (this is what led to the question about mother tongue). Whilst the Russians tend to have reasonable spoken Hebrew, their level of reading/writing is a different manner. At the worst, I will prepare a rough 'Google translate' version of the questionnaire in Russian and then get someone to improve it. 

Probably one can make jokes about the concept of 'user satisfaction' in Russian, although of course, such jokes would be thirty years too late (in those days, one had to be satisfied with what the party supplied, with no option of being dissatisfied). 'Learning style' would also be an interesting concept to have measured in those days....

I obtained a list of companies which rent premises in the industrial area where I normally work; I then sent a letter to them asking if anyone uses Priority. One company (with which I am not familiar) replied and I am trying to arrange an initial interview. This is a tedious business as my correspondent seems to be very suspicious. Hopefully I will be able to conduct the interview in the next few days. This is still part of the pilot process and not final research. 

If things go well, then I will ask to distribute the full questionnaire as it now stands; I will be interested in meta-data - how long it took to complete the questionnaire, whether there are still questions which are not clear and whether there are missing options - but this time I will be more interested in the answers themselves. I will also be checking the 'company' questionnaire.

Tuesday, January 13, 2015

Split questionnaire design

I wrote about finding a paper yesterday which describes the split questionnaire design; there were several papers which I could have chosen but the one that I chose seemed to be fairly representative and understandable. Yesterday I read only the opening page; my eyes glazed over when the paper started discussing heavy statistical analyses.

Today I continued reading the paper: I skipped the theoretical introduction and moved on to the actual experiment which the authors performed. Serendipity strikes its head: they used a questionnaire consisting of 65 questions grouped into 9 blocks; I had 68 questions grouped into 9 blocks (or sections, as I called them). There are two possible algorithms for splitting the questionnaire: 'between block' designs and 'within block' designs.

As I understand the above, I have chosen a 'between block' design: each questionnaire contains complete blocks. A 'within block' design means that each questionnaire contains some questions from each block. As it happens, I was discussing this with the occupational psychologist this morning (prior to reading the paper): she suggested that I use what the authors call the 'within block' design. 

I prefer the 'between block' design as I achieve a higher degree of accuracy for each block: if there are eight questions in a block and each question has five answers, then I can expect a total ranging between 8 and 40. If there are only four questions, then the total can range between 4 and 20, which is obviously a 50% reduction in accuracy. If there is a block with only three questions (and I have such a block), then reducing the number of questions will seriously damage the accuracy.

On the other hand, the authors write: feelings of boredom [may] primarily [be] caused by repetition of the relatively similar questions within blocks that measure similar constructs. Boredom occurs less in the within-block SQD because there are less of these similar questions with each block.

Again, prior to reading the paper, I had become aware of something similar to this problem. I had assigned to one questionnaire the blocks about spreadsheet efficacy and learning style, and to the other, blocks about training, ownership and satisfaction. This was done very much ad hoc late last night but afterwards I began to have misgivings: one questionnaire has blocks which are strongly connected with Priority and the other has questions which are totally unconnected. Thus I have decided to revise the questionnaires so that one has the training, ownership and learning style blocks whereas the other has efficacy and satisfaction. I am also going to standardise the order of the blocks within the questionnaire, for my own benefit.

I did not see any mention in the paper of 'compulsory' (common to all) and 'optional' blocks within the questionnaire so this seems to be my personal contribution. 



My understanding has deepened after having read a few more papers on SQD. What I was calling 'compulsory blocks' are normally named 'core components' whereas the 'optional blocks' are non-core components. Apparently, a more normal design is that there exists one questionnaire which is composed of the core components and all the non-core components; each respondent is directed to answer the core components along with randomly assigned non-core components. My twist is to create two questionnaires, each consisting of core and non-core components, where the respondent has to answer all the questions. 

One might say that this will lead to problems when storing the data and then analysing them, but as I will be writing a dedicated program for storing the results and analysing them (or at least, outputting data which will be analysed by a statistical package), I'm not worried about this. I have a high level of database self-efficacy!

Monday, January 12, 2015

Reinventing the wheel

There's nothing like reinventing the wheel to make one feel foolish. On the other hand, to independently hit on an idea should be a sign of intelligence. I am referring to my previous blog when I realised that in order to get responses to all the questions in my research questionnaire, I should divide it into two (or three, or four) separate questionnaires and give each person one of those questionnaires. 

I have discovered that this technique is called 'split questionnaire design' and there are plenty of papers available on Google Scholar which discuss it. Most of them seem very heavy on the statistics, which is something that I can leave for my supervisor. I found one interesting paper (by Adig├╝zel and Wedel) entitled "split questionnaire design for massive surveys", which seems to contain everything that I need.

I notice that the abstract finishes "... as a result of reduced respondent burden, the quality of data using split questionnaire design improves". Amen.

Removing the blinkers (Research questionnaire 6)

I woke up this morning at 4:30am - not in a cold sweat, but with two pressing issues on my mind. One is concerned with my daily job, which need not concern us here, and one is concerned with the pilot study for my doctoral research.

I handed out six questionnaires to people to be reviewed; most were not returned promptly and I had to remind the people. One of them is now in Canada for a week (she apparently answered the questionnaire instead of reviewing it) so I shall have to forget her, and another has yet to respond. But from the few responses which I did get, one thing is very clear: the questionnaire is too long. I see now that I wrote about this before.

For an hour, I twisted in bed, thinking about how I could reduce the number of questions in the questionnaire without affecting its integrity - this is similar to how to pour a quart into a pint pot. I reached no conclusions. After I got up, I took the dog for her morning walk, and on the way, the answer (or at least, an answer) popped into my head. My thinking, as always, was blinkered, and once I removed those blinkers, I could see the solution. As always, this seems obvious in retrospect, but I've never seen this technique used in practice (i.e. in all the research papers which I read).

Create two questionnaires! 

There are questions which have to be in both questionnaires, such as age, etc, as well as 'the dependent variable', the extent of spreadsheet usage with data exported from Priority. But not everyone need answer other sections of the questionnaire: for example, half of the respondents can answer about satisfaction with Priority whereas the other half can answer about cognitive style. As a result, different questions/sections will have a different number of responses, but this shouldn't present a statistical problem as long as there are over 30 answers per question.

I have written to my supervisor asking about this as I feel that this is a critical issue and I don't want to continue without advise. I will also look in the research book that I bought at his suggestion in the summer. I would try and find references to this technique but I don't know what to look for (this frequently seems to be a problem).

Obligatory science fiction reference: the solution above brings to mind the third section of Isaac Asimov's "The Gods themselves". Why should I be satisfied with two questionnaires when maybe I need three or more? My objective is to have at most 40 questions in the questionnaire, and if this means that each respondent will answer one of two, three or four questionnaires, then so be it. There will be no problem as long as the program which records the responses knows which version is being answered.

Sunday, January 04, 2015

Research questionnaire (5)

The version of the questionnaire which I have sent to six reviewers consists of 68 questions in nine sections:
  • Priority usage – 5 questions
  • User training – 4 questions
  • User ownership – 3 questions
  • User satisfaction – 8 questions
  • Learning style – 11 questions
  • Spreadsheet competency – 14 questions
  • Spreadsheet efficacy – 8 questions
  • Spreadsheets/Priority – 11 questions
  • Personal details – 4 questions
The data (questions) are stored in a hierarchy: options belong to questions, questions belong to sections, sections are connected to versions along with their display order. Thus I can easily create a new version of the questionnaire which eliminates a section or two. I can also change the order of the sections without changing the questions asked per section.

I may well have to do this as the feedback which I have received so far is that the questionnaire is too long. The respondent also did not understand the point of the questions about learning style. As the the spreadsheet/Priority section is of very high importance, I will probably change its position so that it comes after user ownership.

If the other reviewers are in agreement about the length of the questionnaire, then I have four options:
  1. Remove sections
  2. Keep all of the sections but remove superfluous questions
  3. Remove questions from sections and delete sections from the version
  4. Do nothing
There are definitely some superfluous questions in the spreadsheet competency section, thus those could be removed without compromising the final results. I don't want to cut the number of questions per section to below six (for those where it is above six) as I lose accuracy by doing so. I don't want to remove sections as each section belongs to a hypothesis in my research, so removing sections would be reducing the breadth of my research. If I had to make a choice, then spreadsheet efficacy would be the first section to go.

I expect to get all the responses back by the end of the week so that I can discuss them with the occupational psychologist. I will then have to make a decision as to what the next version of the pilot questionnaire should contain, then send the new questionnaire to both the original reviewers and also to a few more people. This process will have to be repeated as many times as needed until I am convinced that I have a suitable questionnaire.


Then I will have to send it to a company in order to get real results.

Less pleasant memories from early 1975

The second incident which has remained in memory from early 1975 definitely happened during the first week of the second term. During the winter term, I had begun accompanying my lunch with half a pint of cider; I was feeling slightly left out during lunch with my classmates. At the Christmas party held at the end of the previous term, I had supplemented my cider along with three shots of vodka and pineapple (about the only other alcoholic drink whose taste I like) which had been forced upon me. I left the student union bar in the Elephant and Castle (south London) feeling rather woozy; by the time I got home to Hampstead (via the tube), I was feeling bad enough to bring up a little (but only a little) vomit.

So: first week, second term. At lunch, I ordered my 'traditional' half a pint of cider, had a sip - and promptly fell underneath the table with strong stomach pains. I didn't drink any more cider that day and haven't had an alcoholic drink since (not that I miss it at all). I went to the doctor who prescribed me a barbiturate to calm my stomach. Unfortunately, the barbiturates gave me strong headaches, so the doctor then prescribed another drug (I think this was Distalgesic) which took care of the headaches ... but gave me stomach aches. So I took more barbiturates and then took more distalgesic ... and by the end of the week, I was walking around in a complete haze, like a zombie. 

In a moment of lucidity, I realised what was happening and took the remaining barbiturates back to the pharmacist so that he could destroy them (it never occurred to me to flush them down the toilet). Presumably I then spent a few days 'drying out'. This incident led to multiple appointments over the next few years in an attempt to discover what was wrong with my stomach. I had one barium meal and in 1977 I had an endoscopy (which at the time was a relatively new procedure). The diagnosis was an inflamed area of the duodenum; I was given (one at a time) several new drugs which didn't really help.

The endoscopy is a story in itself: I had been told not to eat the day before. I walked from where I lived to the Royal Free Hospital in Hampstead, which would have been about a mile and a half. I was given 10 mg of Valium which completely knocked me out and then the procedure took place. I woke up at about 4pm and realised that I had a tube going up one nostril and from there into the stomach. A nurse removed the tube then told me that I was free to go home. I couldn't believe what she said as I was barely awake. "Call a taxi", she said. I had never ridden in a taxi in my life and quite possibly didn't even have any money with me. So like the idiot that I was in those days, I decided to walk home.

I got to the top of Hampstead High Street, near the tube station, which would have been about half the way home, when I collapsed in the middle of the street. Straightaway, a passerby came to me, asked what was wrong and stopped a passing car. The driver took me home, we knocked on the door, someone came and took me in, took me to my room, possibly helped me undress ... and then I slept for another day. This also meant that I hadn't eaten for two days so I was ravenous when I awoke.

Fortunately things have changed. I have had a few endoscopies since then and I am pleased to say that the procedures are better. I had a colonoscopy last year and the dosage of relaxant was so low that I was awake during the procedure (although I couldn't feel anything). My wife drove me home (well, I could hardly walk home from Jerusalem, even without the relaxant).

Stomach pains of one kind or another accompanied me for at least twenty years. As a result of the most recent endoscopy (which was maybe ten years ago), I was diagnosed with a hiatus hernia along with reflux gastritis; since then I take a daily dose of omeprazole and (along with a suitable diet) haven't had any stomach pain since. Omeprazole was only invented in 1979 and became available several years after, so it wasn't around when I needed it the most.

I don't recall suffering from headaches during my pre-Israel years, apart from the above episode, although they had definitely become part of my life in the late 1980s, probably as a result of stress. Unfortunately, I still suffer from them, although much less in the past few years.

[SO: 3721; 3,15,36
MPP: 574; 1,2,6]

Saturday, January 03, 2015

Memories of Richard and Linda Thompson from 40 years ago

In the past few days, with thanks to my new wonder shoes, I've managed to increase the distance that I walk in the evening. Yesterday I completed 6km - and that was on a Friday night, when I frequently feel bloated from dinner. Of course, the further I walk, the longer it takes: yesterday took 55 minutes. Walking such distances is dependent on three factors (that I can think of right now, there are probably more): good shoes, stamina and an occupied brain.

On some walks, I solve programming problems and on others, I think about my doctoral studies (for example, the other night I was thinking about the pilot study). When I was working on the literature review during October, I would actively think about everything which I had read and written that day and what needed to be improved. During the 'old' days of the MBA, I would 'revise' whilst walking. This is valuable thinking time, when I am not distracted by anything, and I'm sure that it's one of the contributing reasons to the comparatively fast pace of my doctorate.

Of course, all the time I'm listening to music - or rather, there is music playing in the earphones that I wear. All of the songs are familiar to me but recently I added about ten more hours of music which I have listened to properly in the past years: Joni Mitchell, Kate Bush, Randy Newman and Richard Thompson.

Which brings us neatly to memories of 40 years ago: January 1975. I was starting my second term of the first year at university and there are two series of extra-curricular events which have stayed in my mind and have had no small importance of my life. At the beginning of this term, I started helping out with the university newspaper, which came out once a fortnight. At first I helped collate the papers but then I started writing. As I recall, my first review was about an exhibition of painting which I saw in the Tate Gallery by Paul Klee.

An added bonus for writing these reviews was that I was able to hear new music (cue Hatfield and the North) and go to concerts for free. I have just proved to myself that memory can be elastic as an event which I thought took place in February 1975 actually took place two months later, but I'm going to treat it as if it took place when I thought it did.

This event is concerned with Richard Thompson, with his then wife, Linda. I have written before about purchasing his first solo album, 'Henry the Human Fly'. In February 1975, I bought his third (and their second) album, 'Hokey Pokey' from a short lived record shop in West Hampstead. At the time, I thought that this record was the bees' knees as it was filled with good songs and good arrangements, but later I revised my opinion in favour of the starker albums which came directly before and after HP.

For some reason, I had missed the preceding record, 'I want to see the bright lights tonight'. The story is that it had been recorded in 1973 but wasn't released because of the Yom Kippur war which led to an oil shortage which led to problems manufacturing records. It was finally released some time in 1974 when I wasn't in Britain, but strangely I didn't buy it when I came home. Maybe it wasn't in the shops. As it happens, I have a memory of visiting Bath University around May/June 1973 with a schoolfriend and hearing the title song on the radio, so it was 'in the can' for some time before being released.

Back to the story at hand. On 25 April 1975 (I thought that this was earlier), R&L had a 'coming out' concert at the Queen Elizabeth Hall in the South Bank complex. I was able to score free tickets to this (as I was reviewing it) and took my girlfriend, G, who wasn't too aware of the Thompsons. On the steps outside the hall there were people selling what I took to be programmes but were in fact copies of a new music magazine called 'Licorice'; I bought one and greatly enjoyed reading it.

The first half of the concert was Richard and Linda acoustic. Someone sent me years ago either a cassette of this  (which I must dig out and digitise) or a cd. I remember thinking during 'Valerio' how Richard was an excellent electric guitarist and it wasn't fair that he was also an excellent acoustic guitarist - leave some room for someone else!

The 'electric' second half, with John Kirkpatrick (accordion) and the ace rhythm section of Daves Pegg (bass) and Mattacks (drums) was as good as the first half; the strange combination of electric guitar and accordion sounded as if it were creating a new kind of soul music.

My records show that the next day I went out and bought a copy of 'I want to see the bright lights tonight'. I discovered shortly afterwards that G had done exactly the same thing. I have a memory of staying overnight at her house once and playing the record the next morning. They had an acoustic guitar (with which I accompanied some of the songs) and a piano, which one of G's younger sisters was learning to play. I wonder whether I tried teaching her one of the songs.

Whilst listening to the songs from IWTSTBLT last night, I wondered whether G updated her vinyl copy to CD, whether she ever listens to it and whether she thinks of me when doing so. I certainly think of her when listening to them.

Friday, January 02, 2015

Painting with numbers

I discovered the website of Randall Bolten a few weeks ago; he wrote a book entitled 'Painting with numbers' and here I will present some of the promotional material for it ....

Bolten presents a book for all who communicate with numbers. "Painting with Numbers: Presenting Financials and Other Numbers So People Will Understand You"will help you to communicate financials and other numbers clearly, understandably, concisely, and – most of all – effectively.

Like any communication skill, how well you can present numbers has a huge effect on how well your audience understands you. When presenting numbers, you are sending signals to your audience about your intelligence, professionalism, and respect for your audience.

You should read Painting with Numbers if you:
  • Communicate with numbers as a central part of your job, as an accountant or other finance professional 
  • Present numbers as part of your function as a lawyer, investor, HR professional, fundraiser, etc. 
  • Work in a fiscal management, taxation, healthcare and other areas of public policy where presenting and understanding numbers are critical 
  • Are pursuing a degree in business, engineering or other career where communicating with numbers is critical 
  • Read, listen to, and make decisions based on numbers presented to you Communicate with numbers only occasionally, but when that skill is needed, the stakes are very high for you 
Many books have been written about writing and speaking, such as Strunk & White’s The Elements of Style and Lynn Truss’ Eats, Shoots & Leaves. But "Painting with Numbers" is the first book about communicating with numbers.

The other day I saw this graph in the newspaper

I sent the graph to Mr Bolten along with the following message: The enclosed graphic was displayed in an Israeli newspaper this morning which discusses how Israelis use the internet. What caught my eye was the fact that there is absolutely no relationship between the length of the histogram bars and the percentage which they are displaying. The bars from top to bottom read
  • Met partners on the internet – 32%
  • Members of family groups in WhatsApp – 70%
  • Visit a doctor after reading about their medical problem on internet – 56%
  • Watch tv series and films via the internet – 63%
  • Listen to music on the net – 80%
He replied: Many thanks for your interesting example! (And thanks for the translation – my Hebrew is ridiculously rusty, and impossible without the vowels.) 

I spent a few minutes myself trying to guess what might have gone wrong with the graph, and I could identify no errors or transpositions that might explain things. I even speculated that the 80%/70%/63%/56%/32% relationship might be explained by the space remaining to the right of the bars, since Hebrew reads right-to-left, but that didn’t work, either. 

The only explanation I could come up with was that the length of the bars was determined by the length of the captions – note that the type size is the same for all five captions, and the blank space on either side of the numbers seems to be about the same in each bar. In other words, this “graph” was not produced in Excel or any other graphing package – it was nothing more than artwork created to make it look like a graph. (This is an interesting variation on Deadly Sin #2, which you will find in the attached “The Deadly Sins of Quantation,” a one-page handout that is marketing collateral for my book, Painting with Numbers.) 

If you ever do find out how & why this graph was generated, please let me know.  It is one of the worst examples of numbers presentation I’ve ever seen, and I may want to write a post about it. 

Thursday, January 01, 2015

Green lighting the pilot study

After completing the literature review for my doctoral thesis, I wrote a new document called the literature synthesis. It wasn't particularly clear what was needed for this stage, so I had a Skype conversation with my supervisor to clarify matters. I understood that the synthesis should basically contain the conclusions from the literature review. As I had already included a fair number of conclusions in the review, I withdrew them from this document and moved them to the synthesis. I then added material about aims, objectives and hypotheses: this material was taken almost directly from the research proposal although it was updated whenever necessary. Finally I added two pages describing the design of the pilot study. This new material was sent to my supervisor about a week and a half ago, and he gave his approval on Monday (the delay was due to holidays).

So now I am starting with my pilot study. This will have two stages: a preparatory stage and an exploratory stage. This is what I wrote about the preparatory stage:
Factors which have to be checked with regard to the questionnaire are:
  • Length: is the questionnaire too long?
  • Flow: do the several sections which comprise the questionnaire fit together?
  • Understandability: are the questions ambiguous?
  • Answers: have all the possible options been included?
As the research itself will be taking place in Israel, the questionnaire has to be translated into Hebrew. This translation will be checked to ensure that it is both grammatically correct and also faithful to the original. The wording may have to be changed in order to enhance the understandability.

This preparatory stage will actually be divided into two: I have given the questionnaire to six people (two psychologists, two ERP users from my company, one ERP user from another company and one university lecturer) for criticism; on the basis of their input, I will update the questionnaire and distribute it to a wider population. In both cases, the people receiving the questionnaire are requested to criticise it and not to answer it.

Assuming that the questionnaire's format is finalised after the second round of the preparatory stage, I will then distribute the questionnaire to users of a company which uses Priority (I have to choose a company and also obtain their permission). This is the exploratory phase, and the respondents will be requested to answer the questionnaire. This is an important stage as it tests the questionnaire on a real population. The answers are also important for it may be that questions will have to be removed: there's no point in having a question which asks for an opinion if everyone answers in the same manner. The results of this phase may be used as part of the final research, but this depends on how many changes are made to the questionnaire after the first field test. It may be necessary to run a second field test if major changes are made, but the whole idea of the two preparatory tests is to freeze the questionnaire's format before it goes out into the field.

I imagine that the pilot study should take about five-six weeks to complete. Once this is done, I can write up the results of the pilot study and submit everything which I have written since the research proposal was accepted to the research committee. As my textbook (IBR2) states,
The DBA Research Committee critically evaluates the literature review submission and decides whether or not it is of an acceptable standard to justify progression to the final stage.
The Committee looks for evidence that all current relevant literature has been identified, read, critically reviewed and synthesised to act as a basis for the development of the research aims and objectives. The Committee also looks for evidence that the proposed research methodology has been developed directly from this synthesis. The literature review submission must demonstrate a clear progression from literature base to current research. The student has to use the literature review submission to demonstrate that the current research a) has been logically and systematically developed from the existing knowledge (literature) base and b) will contribute directly to that knowledge base.

I can devote some time, while the pilot study is running, to integrating the various sections that I have written in order to make a more coherent document (not that much work should be needed). I think that the introduction needs to be improved. This submission will actually be the first half of the final thesis so it is important that it be written well - although of course, I will be able to edit the text before the final thesis is submitted.

This is an important milestone! Next time I'll write about the questionnaire itself.