7.4 IN CONCLUSION

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 
119 120 121 122 123 124 125 126 127 128 129 130 131 

Content-analysis procedure, and particularly, the reliability check, are very

time-consuming. Do we really have to follow them?

Well, what can I say? If accuracy and reliability matter to you, then the sane

answer has to be ‘yes’, especially if, in the case of a dissertation based on

repertory grid technique, your work is to be judged against scholarly

standards by other people. And if accuracy and reliability don’t matter to

you, should you really be using repertory grid technique in understanding

other people?

Finally, I should mention that I have not described one very interesting

bootstrapped content-analysis technique that aggregates individual interviewees’

grids while capturing information about the structure of their

Table 7.9 Content-analysis procedure: factors related to book sales as seen by publisher’s sales staff

Category Constructs

No.,

%

%

Similarity

H-I-L

value

Pricing decisions 1.2 Affordable – less affordable at that size 92 H

3.1 Easily produced – difficult to produce at a reasonable price 83 H

4.4 For the under-Ј20 market – will sell at Ј25 100 H

5.5 Easy to price – standard of illustration may make it too expensive 10, 83 I

6.2 Looks like good value – looks too cheap to be worth the price 20 75 I

6.3 Under Ј20 and soft cover only – would need to sell for Ј35 92 H

7.1 Cheap and cheerful – only libraries could afford it 100 H

7.2 Well advertised, could sell at that price – can’t produce it at the price

that would reach the buyer

92 H

7.3 Affordable illustrative materials – requires a level of illustration we

couldn’t provide at the price

100 H

7.7 Would sell at a higher price than normal for its market – would its

market buy something at the low price required?

58 L

Buyer 1.1 Well suited to the younger market – not our sort of customer 100 H

characteristics 1.6 Appeal to intelligent lay reader – appeal to specialists 75 I

2.5 Pitched just right at readers – presentation too ‘‘pop’’ 67 I

3.7 A healthy undergraduate market – more postgraduate 9, 83 H

4.6 Attractive, our sort of market – aimed at a more conservative sort of

customer?

18 92 H

5.7 A good standard coverage – too narrowly focused for our market 83 I

6.1 Conventional cover, buyers will accept – ring-bound covers: buyers

don’t like

50 L

6.4 Deals with the topics our market likes – some expected topics missing 92 H

6.6 Our sort of book – not sure if it’s for our readers 58 L

Competitors 1.3 No current authors competing – several other competitors 90 H

2.1 Will take 2 years for this edition to be superseded – not aware of

immediate competitors’ plans

67 I

2.4 No similar volume currently in print – several current competitors 75 H

3.3 Competition outdated – competition strong 8, 50 L

4.2 New edition needed urgently to compete – no new editions planned by

competitors

16 83 I

5.1 Can sell more of this book than other publishers can of theirs – can’t 75 L

6.5 No competition – lots of competition 92 H

7.9 Cheaper than our nearest competitor – competitors’ pricing policy

enables them to compete effectively

66 L

Promotion 1.7 Aimed at the trade – general advertising 66 L

2.2 Worth an extensive advertising expenditure – modest budget only 33 L

3.4 Will do well in long run without expensive advertising – will need

constant support to do well

83 H

4.1 Big advertising budget – small advertising budget 8, 83 I

5.4 Modest advertising via direct mailing – extensive advertising using all

promotional means

16 100 H

5.6 Some conference promotions – trade promotions only 83 I

6.7 Advertised heavily in trade press – not advertised in trade press 75 I

7.4 Will sell at this price without promotion – not worth promoting

extensively, will sell anyway

70 I

Popularity 1.4 Likely to sell – low volume 92 H

of topic 2.3 Demand likely to be high – little demand 67 I

3.2 Current demand – demand well satisfied by others 6, 67 I

3.5 Will walk off the shelf – slow sales likely 12 83 H

5.3 Will sell well – won’t sell 100 H

5.8 We’re doing well with this topic – our firm doesn’t do well with this

topic

75 L

Contents 3.6 Good basic read – too much like a reference text to compete well 67 I

4.3 Covers the ground well – essential bits missing 92 H

5.2 A good standard coverage – too narrowly focused for our market 6, 75 L

5.9 Standard topics there as expected – missing essential contents 12 100 H

7.6 Reads well – difficult to read 66 L

7.8 Coverage ensures its likely to become a standard – flavour of the month 92 H

Design 1.4 Illustrations well laid out – cramped appearance 75 M

4.5 Nicely packaged – likely to fall apart 3, 70 L

7.5 Well proportioned – wrong size for this sort of book 6 66 L

Totals 50,

99

construct system about a given topic in a way that includes details of personal

values, and about the implicational connections between constructs at several

levels of detail. Hill’s content-analysis technique (Hill, 1995) depends on an

understanding of constructs as values, a topic which I wish to examine in

Chapter 8, and you might like to read that chapter first before turning to Hill.

THINGS TO DO

It’s difficult to practise a full content analysis by means of a worked exercise,

since you need lots of data and several grids to do anything useful. The

following exercises focus attention on different stages of the process.

Exercise 7.1 Identifying Categories

(a) Take a look at these 19 constructs, a very small subset of comments

overheard at a wine tasting.

Consistent quality – quality inconsistent

Smooth – petillant

Unreliable – always reliable

A long finish – little if any finish

Fruity – grassy

Musty and stale – fresh and bright

Expensive – cheap

Needs to rest and air – drinkable straight on opening

Scented and flowery – deep and heavy

Cloudy – clear

Deep colour – colour rather shallow

Yeasty – clear of yeast

Chocolate overtones – citrus overtones

Heady – light

Sweet – dry

Ready for immediate drinking – will benefit from laying down

Old and brown – young and fresh

Robust with tannin – gentle, without tannin roughness

Overpriced – a bargain.

(b) Devise a simple scheme of four or five categories, and then allocate the

constructs to those categories.

(c) Find a friend who claims to know a lot about wine, and ask him or her to

devise their own set of categories, allocating the constructs to them.

(d) Now place the two data sets into a single reliability table, along the lines

described in Section 7.2.1, step 4.2. Discuss the similarities and differences.

Then recast the table, laying it out so that, however they’re labelled, the

categories which you’re both agreed on run from left to right and top to

bottom of the table. (In other words, turn a table that looks like Table 7.2

into one arranged like Table 7.3.)

(e) Argue until you have agreed a set of common categories, and jot down the

definitions.

(f) Finally, both of you repeat the analysis, independently, but using this

single set of categories. Is the result an improvement over step (d)? (Work

out the percentage similarity score for the ‘before’ and the ‘after’ table and

see.)

I can’t provide you with any answers against which you might check your

working, since I don’t know what categories you and your collaborator might

devise! However, in case you can’t find a collaborator, and want to practise

setting up the reliability table as in step (d), you may get a feel for this stage of

a content analysis if you look at a sample category scheme which I have

provided as Appendix 1.12.

Exercise 7.2 Practising Content Analysis: D-I-Y

Take a small data set of your own, 50 constructs or so, before you first commit

yourself to a full-sized set of 400! Run through the various steps described in

Section 7.2.1. Once you’ve completed them, you’ll recognise that the

procedures are fairly straightforward, and certainly much less confusing

than they might appear at first glance.

When you’ve finished, continue this chapter by

returning to Section 7.2.3.

Exercise 7.3 Preparing Grid Data for Honey’s Technique

Here’s a little refresher in the calculation of sums of differences, checking for

reversals, and working out % similarity scores. As a reminder of what you

need to do in Honey’s content-analysis technique, you’re asked to compute

these against the supplied ‘overall’ construct, and to prepare your grid for

content analysis.

As a senior manager in the oil business, you appreciate the value of experience

in your project managers, but you know that experience isn’t just a matter of

age; some of your younger engineers run their projects just as well as the older

ones. You decide to capture the knowledge and expertise they have that makes

the difference between success and failure in project management. Eight

project managers of varying expertise (as defined by checking the records of

all the projects they’ve managed in the last 10 years) are your elements,

presented anonymously to 20 senior project managers who have managed

major projects in the past themselves, while being responsible for other project

managers at present.

Just one of the grids, from respondent no. 8, is shown below as Figure 7.2. It is

already marked into strips ready for content analysis. Before you do this, you

have to write onto each strip:

(a) The sums of differences between each construct and the ‘overall’

construct.

(b) The corresponding % similarity score (work it out from the formula in

Section 6.1.2, step 7, or, much easier, just look it up in Appendix 4).

(c) Both of the above, reversed.

(d) For each construct, choose the lower of the two values (reversed/

unreversed) and circle it.

(e) Mark each construct with ‘H’, ‘I’, or ‘L’ depending on whether it is in the

highest third, intermediate third, or lowest third of % similarity scores in

this interviewee’s grid.

Appendix 1.13 shows you what your result should look like.

THINGS TO READ

A good way of complementing your knowledge of Honey’s technique is to

read his original paper:

. Honey, P. ‘The repertory grid in action’. Industrial and Commercial Training

1979, 11, 452–459.

Figure 7.2 Exercise with Honey’s technique

And while you’re ordering that item through inter-library loan, why not

order the October and December issues of the journal as well? (That’s

1979, 11, 10 and 1979, 11, 12.) Each contains another article, presented in

the same, user-friendly way, showing a different way of using grid

technique.