2024年5月7日发(作者：压缩文件打不开怎么办)

ACM SIGSoft Software Engineering Notes 18(3): A42-A43 (1993)

CASE and 4GL Product Users’ Participation in

Software Engineering Research

Stephen G. MacDonell

Department of Information Science

A recent graduate student project required access to a number of real-world systems

developed using computer aided software engineering (CASE) tools and/or fourth

generation languages (4GLs) at business sites in the United Kingdom and New

Zealand. Obtaining cooperation for this type of participation, however, was

unexpectedly time-consuming and difficult. This short paper reports the results of the

mail/telephone campaign that was undertaken at the outset of the project when

expectations were high. It is hoped that the experiences reported here may make other

researchers more aware of the problems that can occur in attempting to obtain

software engineering data from this environment. Recommendations concerning the

most useful avenues are also provided.

Introduction

A study considering techniques for the assessment of functional complexity was

begun in October 1990. It was immediately obvious that the success of the study

would depend almost exclusively on the availability of real-world system and

development data. Of specific interest were specification details, from data models

and data flow diagrams, and project management records. Initial contact by mail

began in November 1990 and continued concurrently with other research tasks until

an often-postponed final deadline of April 1992. The degree of response obtained

from both software development sites and automated tool vendors/distributors is

shown in Tables 1 and 2.

Response Results

From Table 1 it can be seen that a total of 340 distinct development sites were

contacted, in an effort to obtain a large representative sample. The overall rate of

response, however, was just 37%. Particularly disappointing was the low response

rate from UK CASE product users, at just 24% of those contacted. This was in spite

of the fact that the research was ‘free’, it was to take up very little of the participating

organisations’ time and resources, and was hopefully to lead to outcomes beneficial to

those organisations. Furthermore, of the 152 replies that were received, continued

response was only maintained by thirty-nine, leading to a group of twenty-two

granting preliminary agreement and to a final sample of just ten sites.

A number of reasons were given by responding sites as to their unwillingness or

inability to take part in the study. A total of forty sites stated that their automated tool

usage was minimal or had only just begun. Moreover, twenty-two others cited no tool

use; surprisingly, ten of these twenty-two were from CASE product user lists.

Whether this absence of tool use was because these sites had abandoned the tools, or

1

because the organisations were mistakenly on the lists, however, is unclear. King [1]

states that CASE is still relatively new, with widespread use only being achieved in

the last two to three years. This may help to explain the poor response level achieved

and the low usage in those organisations that did respond. Another contributor,

although not explicitly cited in the replies, may have been a lack of success with the

products. Although automated tool use is becoming increasingly accepted [2], success

is unfortunately not inherent with the purchase of automated products. In fact, failures

are relatively common in situations where organisations have purchased a product as

the solution to their development problems, but have failed to address the equally

important issues of effective training, management commitment and the adoption of

appropriate work methods. Organisations, however, that have directed their attention

to these problems as well as to their technical requirements have achieved success

with automated tools. It was a requirement of this study that the participating

organisations were committed and relatively mature product users - this may have

inadvertently precluded a large number of those contacted. Experience from the

United States, however, where successful CASE usage would appear to be increasing

[3], would suggest that this may not be the situation for long.

Number of sites

Number of repeats

- Total letters

Number of replies

Number of calls

- Total responses

- Percent response

Further contact

Preliminary agreement

Final agreement

Minimal use of tool(s)

Just starting with tool(s)

Project data unavailable

Plan off-shelf/third party

Have rejected tool(s)

Receivership/Takeover

Tool(s) for training only

No resources to help

Security/Confidentiality

Still evaluating tool(s)

No CASE/4GL use

Sent on to another site

No reason given

UK CASE

156

69

225

45

9

54

24%

21

13

7

5

10

11

0

0

0

1

4

6

3

10

4

6

Site type:

NZ CASE

80

0

80

33

0

33

41%

6

3

0

10

3

3

5

2

1

1

2

1

0

0

2

3

TOTALS

4GL

21

0

21

9

0

9

42%

0

0

0

1

0

4

0

0

0

0

1

1

0

1

1

3

Unknown

83

3

86

39

17

56

65%

12

6

3

4

7

5

0

0

1

0

2

2

2

11

2

3

340

72

412

126

26

152

37%

39

22

10

20

20

23

5

2

2

2

9

10

5

22

9

15

Table 1: Development site response results

2