Back to Main Page

4.4.6: Princals multivariate analysis of Dependent Variables of the Dingle data.
Of the Dependent Variables listed in Appendix II, 30 variables were used for the Princals analysis. The eigenvalues for the first and second dimensions which were extracted by the Princals analysis were 0.4416 and 0.1618 respectively. Fig. 4.8A shows the object scores of the respondents on the first and second dimensions, from which five groups of respondents were selected. These groups are interpreted on the basis of five variables which were selected with the help of the component loadings of the variables on the first and second dimensions shown in Fig. 4.8B.

The values of these variables for the six groups are shown in Table 4.6. The five variables are as follows: the current fishing method of the respondent (curmet24), assessment of flesh quality of catch (quali372), who currently makes the decisions on board (nowdec84), which non-quotum fish species are regular part of catch (nonquo861) and overall fishing trends during the 1990's (tren115.4).

Table 4.6: Values of five selected variables for respondents grouped on the basis of Princals analysis of 30 Dependent Variables of the Dingle questionaires.

Group	Resp.	CurMet24	quali372	nowdec84	nonquo861	tren1154
	1	boards+pair	top	skipper	none	-
I	10	trips	top	skipper	all	-
	3	ProOwn	top	skipper	prawn,sprat	down
	6	angling	top	skipper	<10m	up
II	25	ProOwn	reas	skipper	wink+salm	down
	30	angling	top	skipper	<10m	down
	5	OwnFisOrg	good	skipper	none	down
	7	boards	good	skipper	none	down
III	14	boards	good	skipper	dogfish	up
	20	boards	top	skip/consult	none	stable
	8	Proc	reas	skip/consult	salm,crab,prawn	down
	2	boards	good	skip/consult	squid+prawn	down
	11	boards	good	skip/consult	none	down
	12	boards	good	skip/consult	none	stable
	13	boards	good	skip/consult	none	down
IV	16	boards	good	skip/consult	skate,prwn,ling	down
	17	boards+pelag	good	skip/consult	prawn	up
	21	boards	good	skip/consult	none	down
	22	boards	good	skip/consult	prawn	stable
	24	gillent	good	skip/consult	dogfish	up
	9	pair	good	skip/consult	salm+lobs+oyst	down
	4	lobspot	top	skipper	lobster	down
	15	lobspot	top	skipper	lobster	down
	18	pots+ferry	top	skip/consult	lobster	down
	19	lobspot	top	skip/consult	crab,salm,lobst,mack	down
	23	lobspot	top	50:50	lobster	down
V	26	lobspot	top	skipper	lobster	down
	27	lobspot	top	skip/consult	lobster	down
	28	pots+drift	top	skip/consult	lobst+salm	down
	29	pots+drift	top	skip/consult	lobst+salm+mussel	down
	31	lobspot	top	50:50	lobster	down

Group I assessed their catch's flesh quality as top, the skipper makes the decisions on board and had no information on fishing trend in the 1990's. One respondent in group I used to fish on trawlers and the other is a tourist operator.
Group II is mixture of two fish processors/boat owners and two angling boat operators. Like group I, the skipper makes the decisions in group II, three of the four assessed their catch as top quality and one as good and non had a quotum. Three of the four respondents in group II had seen the overall trend in fishing go down in the 1990's.
In contrast group III are all involved in trawling, mostly fish for species which have quota and assessed the flesh quality of their catch as good, bar one who assessed his catch as top quality. However, like group I and II, the decisions on board of the group III respondents were mostly taken by the skipper, again except one. Their experience of the overall fishing trends was mixed, with two having seen it go down and two said it was stable or rose.
Group IV is entirely composed of medium sized fishermen using trawlers and trawling, except one who gillnets from a trawler. They all assessed their catch's flesh quality as good, in all cases the skipper consults with the crew before taking decisions on board and most fish for species which have a quotum. Some also fish for species which don't have a quotum, but non do so exclusively. Group IV had a mixed assessment of the fishing trends during the 1990's, with five reporting it as going down and four experienced as it having stayed stable or having gone up.
Group V is made up of the lobster fishermen amongst the respondents, who rate the flesh quality of their catch as top, because the lobster are kept and sold alive. Most consult their crew, if they have them, before making decisions on board and all fish for non-quotum species such as lobsters, crab and other shellfish as well as salmon, and in addition, if one fished with a boat of less than 10 m in length, no quota are needed. All of group V had experienced a downwards trend in fishing during the 1990's.
When one examines Fig. 4.8 and Table 4.1 it is clear that all of the lobster fishermen are in the lower left hand quarter of the plot, in other words they were found to have negative values on both the first and the second dimensions in the Princals analysis. The trawler fishermen were predominantly concentrated in the lower right hand side of the plot and the maritourism operators and the non-fishing owners plotted in the top two and lower right hand quarter of the plot.

4.4.7: Combined Princals and Overals multivariate analyses of Traditionality Variables and Sustainability Variables of the Dingle data.
As discussed in Chapter 1, many instances have been found of traditional and indigenous knowledge systems aiding the sustainable exploitation of natural resources. In order to examine the relationship between the variables which indicate the traditionality of the respondents and the variables which indicate how sustainable the fishing practices of the respondent were, 15 traditionality variables (Fig. 4.9A) were subjected to a Princals analysis first and the object scores then taken as the variables for the first data set for an Overals analysis. The eigenvalues for this Princals analysis were 0.4134 and 0.1541 for the first and second dimensions respectively. A second Princals analysis was done on 15 sustainability variables (Fig. 4.9B) and the object scores from that analysis used as the second data set for the Overals analysis.

The eigenvalues for this second Princals analysis were 0.3085 and 0.1903 for the first and second dimensions respectively. The object scores for the first two Princals dimensions from both the traditionality and sustainability data sets are listed in Table 4.7, these were categorised in the usual manner for the Overals analysis.

Table 4.7: Object scores from two Princals analyses, one on 15 Traditionality Variables and the other on 15 Sustainability Variables, for the respondents from Dingle.

Respondent	Traditionality Analysis			Sustainability Analysis
	Dimension 1	Dimension 2		Dimension 1	Dimension 2
1	-1.5	0.98		-0.45	-2.81
2	0.49	-0.14		-1.43	0.59
3	0.41	0.65		-0.14	0.5
4	-1.5	0.21		2.21	0.46
5	0.34	-0.07		-0.94	0.91
6	-0.38	2.86		-0.89	-0.24
7	1.46	1.65		-1.85	-0.07
8	-1.44	-1.25		0.28	1.11
9	0.95	-0.94		-0.62	0.19
10	-1.2	0.1		-0.5	-1.73
11	1.41	1.19		-0.86	0.4
12	0.47	-0.14		-0.62	-1.63
13	1.53	-1.46		-1	1.41
14	-1.55	0.24		-0.31	-0.09
15	0.17	1.34		0.98	-1.6
16	0.76	0.85		-0.84	-0.44
17	-0.1	-2.15		-0.92	-0.16
18	-1.54	1.01		1.55	0.21
19	0.6	0.12		0.14	1.25
20	0.24	-0.59		-0.32	-1.45
21	-0.49	-0.7		-0.27	1.1
22	1.17	-0.59		-0.06	0.01
23	-0.29	-0.29		1.88	-0.44
24	0.5	-1.44		-0.48	1.48
25	-0.67	-0.41		0.12	0.84
26	-1.56	-0.49		1.2	0.7
27	1	0.02		1.57	-0.21
28	1.36	0.11		0.75	0.06
29	0.46	0.3		-0.21	0.14
30	-0.08	-0.36		0.57	-0.89
31	-1.02	-0.62		1.45	0.41

The reason why the two sets of 15 variables were not used directly in an Overals analysis was that the Overals programme is best suited for testing relationship between sets of data which have much less variables than cases, as discussed in Chapter 3. A plot of the Overals object scores for the Dingle respondents are shown in Fig. 4.10A and the Overals component loadings for the Princals dimensions are shown in Fig.4.10B.

The eigenvalues for the Overals analysis were 0.758 and 0.650 for the first and second dimensions respectively. The Overals variables are the Princals object scores on dimensions one and two for the traditionality data set (Tradit1 and Tradit2) and the Princals object scores on dimensions one and two for the sustainability data set (Sustain1 and Sustain2) as is shown in Fig. 4.10B. Seven groups of respondents were identified (Fig. 4.10A) and these are now discussed on the basis of four traditionality variables (Fig. 4.9A) and five sustainability variables (Fig. 4.9B). The values of these variables for the seven groups are shown in Table 4.8.

Table 4.8: Values of four traditionality and five sustainability variables for respondents grouped on the basis of Overals analysis on the first and second dimension object scores of a Princals analysis of 15 traditionality and 15 sustainability variables from the Dingle questionaires.

Group	Resp.	Traditionality Variables					Sustainability Variables
		ReasFi12	Ancs1710	Toyr2413	NoNa48.2		HighstHP	grdson43	choice58	fisaverg	Practs73
	24	tradit+nowork	3	28	1		670	no	lesfis+overhds	3.2	lessmeshes
	13	family	4	13	1		330	no	asnow	2.7	bigmesh
	4	commun	0	28	3		20	no	lesfis+overhds	1.4	smallob+1claw+reintros
	23	lovesea	3	36	1		15	yes	lesfis+overhds	2.3	Vnonberfems
I	26	likeit	0	20	1		10	no	lesfis+overhds	2.1	stockHarbsmalls
	31	emplfisfact	1	10	1		6	maybe	lesfis+overhds	2.0	noVberrback
			Mean 1	Mean 23.5	Mean1.5		Mean 12.75			Mean 1.9
	8	likeit	0	16	1		0	inprocess	lesfis+overhds	3.2	none
II	18	nowork	0	26	3		30	don'tknow	lesfis+overhds	2.1	none
			Mean 0	Mean 21	Mean 2		Mean 15			Mean 2.7
	14	lovesea	0	15	2		0	yes	morfis+overhds	3.2	leavesmalfis
III	30	nowork+lovesea/work	4	20	4		40	don'tknow	lesfis+overhds	2.0	returnfis
			Mean 2	Mean 17.5	Mean 3		Mean 20			Mean 2.6
	1	lovesea	0	25	4		0	don'tknow	dontknow	3.7	none
IV	10	lovesea	0	1	6		0	yes	morfis+overhds	2.5	retbass+bluesetc
	15	likeit	2	18	2		15	yes	morfis+overhds	2.1	Vsmallalso
			Mean 0.7	Mean 14.7	Mean 4		Mean 5			Mean 2.7
	6	family	2	33	2		210	yes	morfis+overhds	2.6	bighooks
	7	family	4	39	2		1200	yes	morfis+overhds	3.6	bigmesh
V	12	family	3	15	2		450	yes	lesfis+overhds	3.2	leavspawners
	16	family	4	43	2		1200	yes	morfis+overhds	2.3	bigmesh
			Mean 3.25	Mean 32.5	Mean 2		Mean 765			Mean 2.9
	2	family	3	14	3		1200	no	morfis+overhds	3.0	bigmesh
	3	family	3	1	2		0	no	lesfis+overhds	2.2
	5	family	4	6	2		1200	no	lesfis+overhds	2.4	none
	9	family	4	17	1		120	no	lesfis+overhds	2.5	notonweeken
VI	11	family	3	35	3		930	maybe	morfis+overhds	2.6	none
	20	family	3	20	1		450	don'tknow	lesfis+overhds	2.4	none
	22	family	4	16	1		1200	no	lesfis+overhds	2.4	none
	25	knewProcess	2	0	1		600	no	lesfis+overhds	2.5	none
	29	family	2	22	1		10	maybe	morfis+overhds	2.7	none
			Mean 3.1	Mean 14.6	Mean 1.7		Mean 634.4			Mean 2.5
	17	tradit+nowork	4	28	1		450	yes	lesfis+overhds	3.1	bigmesh
	19	commun	3	26	1		130	no	lesfis+overhds	1.3	notonnight
VII	21	nowork	2	13	3		425	maybe	lesfis+overhds	2.3	none
	27	family	3	58	1		10	yes	lesfis+overhds	2.3	noVberrback
	28	tradit+nowork	4	32	1		0	maybe	lesfis+overhds	2.3	lobsbackwards
			Mean 3.2	Mean 31.4	Mean 1.4		Mean 203			Mean 2.3

The four traditionality variables are as follows: reason why the respondent is involved in the fishing industry (reasfi12), number of respondent's ancestors who fished (ancs1710), number of years which the respondents was involved in fishing (toyr2413) and number of traditional navigational methods which the respondent mentioned (nona48.2). The five sustainability variables are as follows: highest horse power which respondent fished with (highsthp), whether respondents thinks that his grandson will have the choice to fish in the manner that he himself is fishing now (grdson43), choice between catching less fish with lower overheads or catching more fish with higher overheads (choice58), the average of the fish stocks abundance ratings (fisaverg) and practices carried out by respondent which help to protect fish stocks but are not required by law (practs73).

Group I is characterised by a medium average number of years in fishing and a low average number of fishing ancestors in the traditionality variables. It also had a low average for the number of traditional navigation methods which the respondents remembered. The reasons for fishing were mixed, consisting of love of the sea, tradition and economic factors. In the sustainability variables, group I has a low average horse power, tending towards a negative view on whether grandsons will be able to fish in the way they are and all would see catching fewer fish with less overheads as being the most advantageous. Group I has the lowest average rating for fish stock abundances of all the groups and all practised non-compulsory conservation practices, all of them for lobster stocks.
Group II is different from group I in that it has no fishing ancestors and a somewhat higher average for the number of traditional navigation methods mentioned. In the sustainability variables it has the joint second highest average score for fish stock abundances and did not practice any non-compulsory conservation methods.
The respondents in group III are in fishing because of love of the sea as well as economic reasons, and have the second lowest average number of years fishing experience and listed the second highest average number of traditional navigation methods. Like groups I and II, they have a low average horse power in the boats, but are mixed on the choice of fishing strategy, one preferring trying to catch more fish even if it means greater overheads and the other being of the opinion that lower overheads are a better strategy. Like group II, the respondents in group III were mixed on their views of the grandson's generation prospects in fishing. Group III mentioned conservation practices for finfish.
The respondents in group IV are in fishing because they like it and have low averages for fishing ancestors as well as their years of fishing experience, but mentioned the highest average number of traditional navigation methods. In the sustainability variables, like the previous three groups, group IV has a low average in horse power but unlike group I has a high average rating for fish stock abundances and, unlike groups I and II, tended towards trying to catch more fish with higher overheads. Group III has a mix for the non-compulsory conservation practices, one said he had none, one related to finfish and one was for lobsters.
Group V has family tradition as the reason for being in fishing, unlike the previous four groups. This group also has the greatest average number of years fishing experience as well as the highest average number of ancestors in fishing. In sustainability variables group V has the highest average horse power, all think that their grandson will be able to fish the way they are currently fishing and this group also has the highest average rating for fish stock abundances. They also all reported using fish conservation practices which are not required by law. Three of the four in group V thought that attempting to catch more fish by creating greater overheads, was a more reliable fishing strategy.
In the traditionality variables the respondents in group VI, like group V, said family tradition was the reason why they were in fishing and also had a high average number of fishing ancestors but, unlike group V, this group has the lowest average number of years fishing experience and also a lower average number of traditional navigation methods. In the sustainability variables group VI, again like group V, has a high average boat engine horse power but was generally negative about the chances of their grandson being able to fish the way they are. Most in group VI did not report any conservation practices which are not required by law, unlike group V. The average score for fish abundance ratings was lower than group V and the majority of respondents in group VI thought that catching less fish with less overheads is better for long term fishing success.
Group VII is characterised by a high average of years of fishing experience and also of numbers of fishing ancestors, similar to group V. But unlike group V and VI, their reasons for fishing were a mixture of traditional and economic reasons. Group VII also mentioned a low average number of traditional navigation methods, similar to group I. In the sustainability variables, the average horse power used by those in group VII was medium, much higher than groups I, II, III and IV but significantly lower than groups V and VI. Like group VI, group VII preferred to catch less fish in combination with having less overheads rather than having more overheads and catching more fish. Both the opinions on their grandson's fishing prospects and non-compulsory conservation practices were mixed in group VII, and their average rating for fish stock abundances was medium. Respondents 13 and 24, had relatively many fishing ancestors but have medium and high numbers of years of fishing experience. They only mentioned one traditional fishing method each and both had tradition as their reason for fishing. Again like group VII, these two respondents had medium horse power but were negative about their grandsons' fishing choices. They also had higher fish abundance ratings than the average for group VII.
Overall respondents 13 and 24 and groups V, VI and VII on the left hand side of the graph had the higher horse powered boat engines, while groups I, II, III and IV on the right of the graph were in the lower horse power range (Fig. 4.10A). The groups on the left and the two single respondents all had average numbers of fishing ancestors of three or more, while the right hand side had lower average numbers of fishing ancestors. The middle and lower left hand side of the graph contains those respondents who gave family tradition as the reason for fishing and groups IV and V in the lower section of the graph predominantly preferred catching more fish even if it meant having more overheads and these were also the groups which were most positive about their grandsons being able to fish the same way as they are.

4.5: Quantitative Questionnaire data from Goeree, The Netherlands.

4.5.1: Overview of the questionnaire results from Goeree.
All the respondents were male as, like in Dingle, commercial fishing is an exclusively male occupation in Goeree. The age distribution of the respondents are shown in Fig. 4.1. Respondents ranged in ages from 21 to 73, with most respondents aged between 30 and 50 years of age. One respondent was single, 28 were married and two were divorced. Eight respondents were residents of Goedereede, three lived in Goedereede/Havenhoofd, seven in Ouddorp and four in Stellendam, which are all on the Head of Goeree (Fig. 2.11). A further three respondents lived in Moerdijk, two in Hellevoetsluis and one each in Oostvoorne, Tholen, Haamstede and Hippolytushoeve which are located outside the island of Goeree-Overflakkee. The numbers of children in the families are shown in Fig. 4.2. Of the 30 families, most had two or three children, with two families having none. One of the Goeree respondents was a parttime angling skipper and the other 30 were fulltime involved in fishing. Of the 31 respondents, 16 beam trawled with large trawlers, nine had the smaller Eurokotters, two were cockle fishermen, two were angling skippers, one was a teacher in the fishery school and one was a retired crew member. As in Dingle, all the respondents had had experience of commercial sea fishing. The nature of the respondents' involvement in the fishing industry at the time of interviewing is listed in Table 4.9.

Table 4.9: List of the nature of the Goeree respondents' involvement in the fishing industry at the time of the interviewing.

	Respondents number		Nature of involvement
	1		large beam trawler
	2		Eurokotter
	3		Eurokotter
	4		large beam trawler
	5		Eurokotter
	6		large beam trawler
	7		large beam trawler
	8		large beam trawler
	9		large beam trawler
	10		Eurokotter
	11		Eurokotter
	12		Eurokotter
	13		large beam trawler
	14		large beam trawler
	15		Eurokotter
	16		angling
	17		large beam trawler
	18		large beam trawler
	19		Eurokotter
	20		angling
	21		retired crew member
	22		fishery teacher
	23		Eurokotter
	24		large beam trawler
	25		large beam trawler
	26		large beam trawler
	27		cockle fishing
	28		large beam trawler
	29		cockle fishing
	30		large beam trawler
	31		large beam trawler

As regards formal education, ten attended the sea fisheries school, 15 had attended weekend courses training radio operators and skippers, two had completed the school for sea navigation, one had attended the school for inland waters navigation and two had attended primary school only. All respondents, either in the past or the present, had at least one of their family members involved in fishing. The majority of the respondents were optimistic about future fish stocks, with 24 having reasonable or full confidence in future fish stocks, one thought they would stay as they are at present and six thought the stocks will decline.
When asked whether they thought that their grandson would have the choice available to him to fish in the manner in which the respondent was currently fishing, 13 respondents thought that they would be able to and ten thought that they would not, with eight not being sure what the future situation in fishing would be.
These data give an outline of the 31 respondents which took part in the questionnaires in Goeree.

4.5.2: Princals multivariate analysis of Predisposing System Variables of the Goeree data.
Of the Predisposing System Variables listed in Appendix II, 29 variables were used for the Princals analysis. The eigenvalues for the first and second dimensions which were extracted by the Princals analysis were 0.3452 and 0.2359 respectively. Fig. 4.11A shows the object scores of the respondents on the first and second dimensions, from which six groups of respondents were selected. These groups were interpreted on the basis of six variables which were selected with the help of the component loadings of the 29 variables on the first and second dimensions shown in Fig. 4.11B.

The values of the six selected variables for the six groups are shown in Table 4.10. The six variables are as follows: respondent's domicile (res1), number of children of respondent (chlds4.1), family members currently in fishing (famfis15), number of ancestors who fished (ancs1710), total years of fishing experience (toyr2413) and number of fishing methods which respondent had experience of (nome2415). These were selected from Fig. 4.11B in a similar manner to that described in Section 4.4.2.

Table 4.10: Values of six selected variables for respondents grouped on the basis of Princals analysis of 29 Predisposing System Variables from the Goeree questionnaires.

group	respond.	1res	chlds4.1	famfis15	ancs1710	toyr2413	nome2415
	11	Moerd	1	uncbroncous	4	19	1
I	12	Moerd	1	uncbroncous	4	11	1
			Mean 1		Mean 4	Mean 15	Mean 1
	23	Haamstede	0	-	1	3	1
	9	oud	5	fabrneph	3	19	3
	13	havenh	3	son+broth	3	28	1
II	15	tholen	3	SoBrNe	3	27.5	1
	21	oud	0	nephews	3	29	3
	22	goe	2	inlaw+neph	3	16	1
			Mean 2.6		Mean 3	Mean 23.9	Mean 1.8
	16	Hellevo	2	cousins	2	10	1
III	20	Hellevo	2	uncle	0	25	1
			Mean 2		Mean 1	Mean 17.5	Mean 1
	1	oud	2	sbc	0	27.5	2
	2	OostVoorne	1	cousins	0	27	3
	3	oud	3	uncle	3	13.5	3
IV	24	oud	2	cousins	2	29	3
	26	oud	2	SoBrUncNe	2	24	2
	27	oud	3	son(s)	0	33	4
	29	Hyppolytushoef	3	son+neph	1	23	5
			Mean 2.3		Mean 1.1	Mean 25.3	Mean 3.1
	4	goe	2	son+broth	3	40	3
	6	havenh	4	SoBrUncNe	3	35	2
	7	goe	3	son+neph	2	38	2
	8	goe	2	SoBrUncNe	2	47.5	2
	10	stel	3	SoBrNe	3	38	3
	17	goe	5	SoBrNe	3	47	2
V	18	goe	9	SoBrNe	3	47	2
	19	stel	2	SoBrNe	2	53	4
	25	stel	3	son+broth	3	36	2
	28	stel	2	son+neph	2	37	2
	30	goe	3	SoBrNe	2	53	3
	31	havenh	8	SoBrUncNe	2	43	2
			Mean 3.8		Mean 2.5	Mean 42.9	Mean 2.4
	5	Moerd	3	son+broth	4	38	3
VI	14	goe	0	brother(s)	4	42	3
			Mean 1.5		Mean 4	Mean 40	Mean 3

Group I was characterised by the lowest average numbers of years in fishing and number of fishing methods which the respondent had experience of, but with a joint highest number of fishing ancestors. This group also had the lowest average number of children in the family, had close family members in fishing and both respondents in this group lived in Moerdijk. In fact the two respondents comprising group I are brothers.
Group II has the second highest average number of children in the family and also the second highest average number of fishing predecessors. The respondents in group II came from a variety of locations and had a mix of close and extended family members who fished, they also had a medium average number of years experience of fishing and a relatively low average number of fishing methods which they had used in their fishing career.
Both respondents which make up group III live in Hellevoetsluis, just to the northeast of the Head of Goeree, have extended family members only who fish and have low averages for numbers of fishing ancestors, total years of fishing experience and numbers of fishing methods. They had a medium average number of children in their family.
Group IV differs from group III in that five of its seven respondents live in Ouddorp. Group IV had experienced the highest average number of fishing methods of all the groups, but had a medium average total number of years fishing experience and a low average number of fishing ancestors. Like groups II and III, group IV had a medium average number of children in the family. Group IV had a mixture of close and extended family members who fished. Group V differs from all groups in that its respondents live in Goedereede, Goedereede/Havenhoofd and in Stellendam (only one respondents who lived in these centres of population is a member of another group), they had the highest average number of children, and the highest average total number of years fishing experience. Also all respondents in group V had close family members who fished. Group V had a medium average number of fishing predecessors and a medium average number of fishing methods which they had used.
Group VI differs from group V in that it has a joint highest number of fishing predecessors with group I, but differs from group I in that it has a high average total number of years fishing experience, more like group V. Group VI also has experience of a high average number of fishing methods. The respondents in group VI lived in Moerdijk, which is a centre of population east along the Haringvliet from Goeree-Overflakkee, had a medium average number of children and close family members who fished. Respondent 23 was separated from the other respondents and was the only one to be unmarried and also the youngest fishing with the smallest trawler. He lived in Haamstede which is on the island to the south of Goeree-Overflakkee.
Overall, the respondents with high positive values on dimension 1, the right hand side of the graph in Fig. 4.11A, had few ancestors who had fished. Groups I, II and III and respondents 23, in the upper right hand side of the plot tended to have experience of fewer fishing methods than the respondents in the lower left hand side of the graph. As mentioned already, group V had the highest average years of experience and number of children in the family and this group is the only one in the lower left hand quarter of the graph, i.e.: with negative values for both dimensions one and two. Respondents residing in Stellendam were exclusively located in the lower left hand side of the graph.

4.5.3: Princals multivariate analysis of Predisposing Attitude Variables of the Goeree data.
As with the analysis of the Dingle quantitative data, a Princals analysis was done on the attitudes variables on fish stocks generated by question 61 (Appendix I) of the Goeree data, in order to reduce these 19 variables so that they would not contribute a disproportionate amount of variation to the analysis of the predisposing attitude variables as a whole, before carrying out the Princals analysis of the predisposing attitudes variables. Fig. 4.12 shows the component loadings labelled with the variable name codes.

The eigenvalues for the first and second dimensions were 0.4371 and 0.2016 respectively. From this analysis four representative variables were selected based on similar criteria as the selection of variables in Section 4.4.2, and these were used in the Princals analysis with the rest of the predisposing attitude variables. These variables are as follows: opinion on current sole stocks (solenow), opinion on current turbot stocks (turbnow), opinion on current shrimp stocks (shrinow) and the average score of the respondents on past and current stocks of the species which were important to them (fisaverg).
Of the Predisposing Attitude Variables listed in Appendix II, 54 variables were used for the Princals analysis. The eigenvalues for the first and second dimensions which were extracted by the Princals analysis were 0.2431 and 0.1415 respectively. Fig. 4.13A shows the object scores of the respondents on the first and second dimensions, from which four groups of respondents were identified. These groups are interpreted on the basis of six variables which were selected with the help of the component loadings of the variables on the first and second dimensions shown in Fig. 4.13B.

The values of these variables for the four groups are shown in Table 4.11. The six variables are as follows: membership of national fishermen's organisations (fisorg23), opinion on whether respondent's grandson will have the option to fish in the manner in which the respondent is fishing now (grdson43), future fishing strategy based on trying to catch more fish and also increasing overheads or catching less fish but also reducing overheads (choice58), average score of past and present fish stocks assessments (fisaverg), opinion on trends of future fish catches (catfu116) and opinion on what should have been the maximum horse power allowed during the last 15 years (pasthp66). These variables were selected on the same basis as the variables in Section 4.4.2.

Table 4.11: Values of six selected variables for respondents grouped on the basis of Princals analysis of 54 Predisposing Attitude Variables from the Goeree questionnaires.

Group	Respon.	fishorg23	grdson43	choice58	fisaverg	catfu116	PastHP66
	11	visserbond	no	lesfis+overhds	3.1	fullconf	1200
	16	none	yes	morfis+overhds	2.8	reasonconf	1000
	20	none	no	morfis+overhds	2.375	less	0
I	23	visserbond	maybe	morfis+overhds	3.5	reasonconf	200
	29	Shelfishccop	yes	morfis+overhds	1.75	reasonconf	300
					Mean 2.6		Mean 375
	27	PO	maybe	lesfis+overhds	2.5	reasonconf	600
	22	visserbond	yes	lesfis+overhds	2.7	reasonconf	1300
II	24	visserbond	yes	lesfis+overhds	1.9	less	1500
					Mean 2.26		Mean 1400
	1	visserbond	yes	lesfis+overhds	3.4	reasonconf	1500
	2	visserbond	maybe	lesfis+overhds	3.6	less	300
	4	federatie	yes	lesfis+overhds	2.9	reasonconf	1500
	6	federatie	maybe	lesfis+overhds	2.5	less	1750
	7	visserbond	yes	lesfis+overhds	3.4	less	1000
	8	federatie	yes	lesfis+overhds	3.125	reasonconf	1500
	13	federatie	yes	lesfis+overhds	3.125	less	2000
III	14	federatie	don'tknow	lesfis+overhds	3.625	fullconf	1250
	21	visserbond	yes	lesfis+overhds	2.8	fullconf	1000
	25	visserbond	yes	lesfis+overhds	3.25	fullconf	1500
	26	federatie	no	lesfis+overhds	3	reasonconf	2000
	28	visserbond	yes	lesfis+overhds	3.25	fullconf	1800
	30	visserbond	don'tknow	lesfis+overhds	3.125	reasonconf	1200
	31	visserbond	maybe	lesfis+overhds	2.9	reasonconf	1500
					Mean 3.1		Mean 1414
	3	visserbond	no	lesfis+overhds	3.3	fullconf	1350
	5	visserbond	yes	lesfis+overhds	2.9	fullconf	1000
	9	visserbond	no	lesfis+overhds	3.1	fullconf	1500
	10	federatie	maybe	morfis+overhds	3.7	stable	1500
IV	12	visserbond	no	morfis+overhds	3.3	reasonconf	300
	15	visserbond	no	lesfis+overhds	3.2	reasonconf	2000
	19	visserbond	no	lesfis+overhds	3.3125	fullconf	1200
	17	visserbond	no	morfis+overhds	3.0625	fullconf	1800
	18	visserbond	no	lesfis+overhds	3.0625	fullconf	1500
					Mean 3.2		Mean 1350

The respondents in group I all thought that catching more fish with more overheads was the best future strategy for their fishing enterprise and this group also had the lowest suggested mean horse power for the last 15 years. The respondents in this group had mixed opinions on whether their grandson would have the option of fishing using present-day fishing methods and were varied in terms of fishermen's organisation membership. Three of the four respondents in group I were reasonably confident in future fish catches.
Group II had the lowest average fish stocks rating and the second highest suggested horse power for the last 15 years. Both respondents in group II were members of the Vissersbond and thought that their grandson would be able to fish like they were. They also were of the opinion that the best strategy for their future enterprises was to reduce overheads even if this also meant reduced fish catches, but different in their view of future fish catches, one thinking that it would be reasonable and the other that they would be less.
Group III had the highest average for the horse power that it suggested would have been best during the last 15 years and all thought that a strategy of reduced catches but with less overheads would be best for the future. The respondents in group III were members of both large Dutch national fishermen's organisations, the Vissersbond and the Fisherman's Federatie. Group III also had the second highest average fish stocks rating and varied between thinking that their grandson would have the choice of fishing in the same way they were at present and not being sure, with only one respondent in this group who was of the opinion that this would probably not be the case. While groups II and III suggested similarly high average horse power limits which would have been best for the last 15 years, group II had a lower average fish stocks rating. Group III's opinions on future fish catches varied between full confidence, reasonable confidence and thinking that catches will be less than are now.
Group IV had the highest average score for fish stocks ratings. This group had a somewhat lower average for the suggested horse power limit for the last 15 years than group III and all respondent bar one are members of the Vissersbond. This is unlike group III, who judged the best horse power for the last 15 years to be somewhat higher and nearly half the respondents of this group were members of the Federatie. Group IV, again unlike group III, had the predominating view that their grandsons would not be able to fish as they are at present. A greater proportion of the respondents in Group IV had full confidence in future fish catches than in group III, in fact none thought that catches would be less. Group IV had a mix of opinions on whether it would be best to catch more fish with higher overheads or less fish with lower overheads.
Overall none of the respondents who thought that catching more fish with more overheads would be best for their future fishing enterprises were located in the bottom half of the graph (Fig. 4.13A), in other words those who had this opinion all had positive values on the second dimension.

4.5.4: Princals multivariate analysis of Predisposing Enabling Variables of the Goeree data.
Of the Predisposing Enabling Variables listed in Appendix II, 12 variables were used for the Princals analysis. The eigenvalues for the first and second dimensions which were extracted by the Princals analysis were 0.3889 and 0.2807 respectively. Fig. 4.14A shows the object scores of the respondents on the first and second dimensions. As is clear from this graph, 22 of the 31 respondents grouped closely together in the lower left hand corner. This close grouping reflects the similarity of these respondents, based on the 12 predisposing enabling variables. However, in order to further examine the variability within this group, four variables were selected from the 12 predisposing enabling variables, in the manner outlined in Section 4.4.2, from the graph in Fig. 4.14B.

A Princals analysis was carried out on the four selected variables and the eigenvalues for the first and second dimensions which were extracted by this analysis were 0.4861 and 0.2906 respectively. The four selected predisposing enabling variables were as follows: the current horse power used by the respondent (highhp), the specific problems associated with the current fishing method (probs40), which modern fish finding equipment is used (modfis50) and did respondents have enough catch quota during the 1990's (qo84/90.2). Fig. 4.15A shows the object scores of the respondents on the first and second dimensions, from which eight groups of respondents were selected. These groups are interpreted on the basis of the four variables used in the analysis, the component loadings on the first and second dimensions of which are shown in Fig. 4.15B.

The values of these variables for the eight groups are shown in Table 4.12.

Table 4.12: Values of four selected Predisposing Enabling Variables from Goeree on the basis of which respondents were grouped using Princals analysis.

Group	Resps	HighHP	probs40	modfis50	qo84/90.2
	7	2000	energy	none	moreenough
	14	3600	none	none	justenough
	26	2000	wear	none	justenough
I	28	2400	wear+meshsize	none	notenough
	30	2000	energwear	none	justenough
	31	2000	wear	none	moreenough
		Mean 2333
	23	225	overfish	none	-
II	27	800	seaweed	none	-
	29	150	work+selllicens	none	moreenough
		Mean 391.7
	16	370	helpanglers	fishfind	-
III	20	300	overfish	fishfind	-
		Mean 335
	8	1700	fiscapac	fishfind	moreenough
IV	24	1800	fewHerr	fishfind	justenough
		Mean 1750
	5	300	none	experience	moreenough
V	9	1800	wear	exp+sonar	moreenough
		Mean 1050
	1	1800	none	fishfind	justenough
	2	300	energy	fishfind	justenough
	4	1800	none	fishfind	notenough
VI	11	300	none	fishfind	notenough
	12	300	wear	fishfind	moreenough
	19	300	wear	fishfind	moreenough
	22	-	energy	depthsound	justenough
		Mean 800
	3	300	none	none	notenough
	10	300	wear	none	notenough
	15	300	energy	none	notenough
VII	17	1800	wear	none	moreenough
	18	1800	energy	none	moreenough
	21	-		none	moreenough
	25	1800	none	none	justenough
		Mean 1050
	6	2700	none	fishfind	justenough
VIII	13	2000	none	fishfind	notenough
		Mean 2350

Group I and group VIII had the highest average engine horse power of all the groups. However, in group I all respondents bar one named energy use and wear and tear the main problems associated with beam trawling, while group VIII identified no problems. Group I also said that fish finders were not used for locating fish while group VIII did use fish finders. All respondents except one in group I had enough quota during the 1990's and in group VIII on said he had and said he did not.
Groups II and III are similar in that they both have low mean engine horse powers, named conservation and practical problems as the main problems of their fishing method and, bar one respondent in group II, did not need quota for their method of fishing. However, they differ in that group II did not use the fish finder to locate fish while group III said they did.
Groups IV and VI both used the fish finder in locating fish but differed on the other variables. Group IV had a higher average engine power, named fish stocks conservation as their main problems while group VI varied on this variable, with four naming energy and wear and tear as their main problems and three said there were no specific problems associated with their fishing method. Two of the respondents in group VI reported not having enough quota during the 1990's, while the other six said that they did.
Group V had a medium average engine horse power, had more than enough quota during the 1990's and used their experience to locate good fishing grounds. One of the two respondents in group V said there were no specific problems associated with beam trawling while the others named wear and tear of the fishing gear as a specific problem.
Group VII had a medium average engine horse power and, like group VI, named a mix of wear and tear, high energy demand and no problems as snags associated with beam trawling. The respondents in both these groups also varied on having had enough quota during the 1990's. However, unlike group VI, group VII said they did not use any modern fishing finding equipment to locate fish.
Overall all respondents who reported not having enough quota during the 1990's were on the left of the Y axis of the graph, in other words they all had negative values on the first dimension. Those respondents for whom quota were not applicable all had values of greater than +2 on the first dimension. All respondents with engine horse powers of 2000 hp or greater had values of less than -0.5 on the first dimension and greater than -0.2 on the second dimension. All respondents, except respondents 16 and 20 in group III, with positive values on dimension two, or located above the X axis, said they did not use modern fish finding equipment to locate fish.

4.5.5: Princals multivariate analysis of Intervening Variables of the Goeree data.
Of the Intervening Variables listed in Appendix II, eight variables were used for the Princals analysis. The eigenvalues for the first and second dimensions which were extracted by the Princals analysis were 0.3801 and 0.2591 respectively. Fig. 4.16A shows the object scores of the respondents on the first and second dimensions, from which six groups of respondents were selected. These groups are interpreted on the basis of four variables which were selected with the help of the component loadings of the eight variables on the first and second dimensions shown in Fig. 4.16B.

The values of the four selected variables for the six groups are shown in Table 4.13. The four variables are as follows: traditions in Goeree which have been important in the development of fishing (trads44), experience of overall effect of the EU on sea fisheries (effeu100), experience of fisheries research (rsrch105) and reasons for effect of the black market (why113). These were selected from Fig. 4.16B in a similar manner as described in Section 4.4.2.

Table 4.13: Values of four selected variables for respondents grouped on the basis of Princals analysis of eight Intervening Variables from the Goeree questionnaires.

Group	Respons.	trads44	effeu100	rsrch105	why113
	2	history	noeffect	good	raisedprice
I	23	history	dontknow	reasidea	littlenoefect
	29	history	noeffect	good	-
	4	wanttofis	unfavour	guesswork	raisedprice
II	27	history	unfavour	good+consult	openmarket
	1	lijdzaam	unfavour	good+consult	depresprice
	5	history	unfavour	good	depresprice
	6	sharesyst	reasoneff	reasidea	blakcash+Pricdown
	7	shar+hist	reasoneff	reasidea	depresprice
	8	history	reasoneff	gues+consult	depresprice
	9	maintenance	reasoneff	politic+consult	depresprice
III	12	history	noeffect	reasidea	depresprice
	15	history	reasoneff	reasidea	depresprice
	21	history	verygood	reasidea	depresprice
	22	view+persist	verygood	reasidea	depresprice
	24	history	reasoneff	reasidea	depresprice
	30	view+persist	reasoneff	reason+consult	depresprice
	11	maintenance	reasoneff	survey?	depresprice
	13	shar+reinvest	reasoneff	influpaymnt	depresprice
IV	14	shipbuilddev	reasoneff	politic+consult	depresprice
	25	hist+hardwrk	reasoneff	notaclue	depresprice
	26	comunbackng	reasoneff	survey?	depresprice
	3	dontknow	unfavour	guesswork	raisedprice
	17	BankSnijdHist	unfavour	guesswork	stress+blackcas
V	18	BankSnijdHist	unfavour	gues+consult	stress+blackcas
	19	school	unfavour	guesswork	depresprice
	28	bank	unfavour	influetohi	depresprice
	10	BankSnijdHist	noeffect	guesswork	raisedprice
VI	16	none	dontknow	guesswork	littlenoefect
	31	Mr.Snijder	reasoneff	notaclue	littlenoefect
	20	none	-	-	-

Group I differs from group II in that the respondents in group I saw the EU's effect as not noticeable while group II experienced it as negative. All three respondents in group I saw local tradition as having been important in the development of fishing in Goeree and thought that fisheries research had been good. Reported effects of the black market varied between that it raised prices, that it had little effect and no comment.
Group II indicated local tradition and love of fishing as reason for the development of fishing in Goeree, one respondent thought fisheries research was good, but the other had come to the conclusion that fisheries research merely depended on guesswork for estimating fish stocks. Like one respondent in group I, one respondent in group II also reported that the black market had raised prices while the other respondent in this group was a cockle fisherman and for cockles there is an open market.
Group III was characterised by fishing tradition, personal initiative and the share system of crew payment being given as the reasons which had been important in the development of fishing in Goeree. The respondents in group III also predominantly reported having found the effect of the EU on fishing as positive, with just two of the 12 respondents in this group having found it negative and one not having noticed any influence of the EU.
Group IV found that the black market had depressed fish prices. With respect to the three last mentioned variables, group III is similar to group IV, but, in contrast to group III, group IV had formed a negative view of the fisheries research service.
Group V reported mixed reactions to effect of the black market, varying from that it had raised prices and having caused heart attacks due to stress amongst the fishermen and causing further problems when the black money resulting from it had to be dealt with, to having depressed prices.
As group IV, group V had a negative view of the work of the fisheries research service, having the impression that fish stock estimates were not reliable and resulted from guesswork. One respondent in group V was of the opinion that the fisheries research service had too great an influence on fisheries management and policy making. Group V differed from group IV as it had found the effect of the EU on fisheries unfavourable and saw local institutional support of the fishing industry by the banks and schools, one teacher in particular, as the main reasons why fishing had developed on Goeree.
Two of the three respondents in group VI also thought that fishing on Goeree had benefited greatly from its schools and one particular teacher who had a great interest in the fishing industry and organised weekend courses for fishermen. The third respondents did not think there were any particular factors on Goeree which had had a big effect on the fisheries. Experience in group VI with regard to the effect of the EU varied from none, in the case of two respondents, to positive, in the case of the third respondent, unlike group V. However as group V, group VI had found the fisheries research service to not really know what the state of fish stocks were and they had also gained the impression that their work was based on guesses. Again unlike group V, respondents in group VI reported the effect of the black market to have raised fish prices as well as not to have had any effect. Respondent 20 was atypical in that he said that he had no opinion on three of the four selected variables, and thought that there was no particular factor in Goeree which had helped to develop fishing there. Respondent 20 was not included in any group.
Overall all but one of the respondents who plotted in the right side of the Y axis, or had positive values on dimension one, thought that the black market had depressed prices and the one respondent who did not mention this, had found it to be very stressful and to result in problems with black money. All the respondents below the X axis, or negative values on dimension two, did not think that the fisheries research service did a very good job and had formed the impression that fish stock estimation was based on guesswork. All respondents with values of smaller than -0.9 on dimension one had experienced the effect of the EU on fisheries either as neutral or positive.

4.5.6: Princals multivariate analysis of Dependent Variables of the Goeree data.
Of the Dependent Variables listed in Appendix II, 28 variables were used for the Princals analysis. The eigenvalues for the first and second dimensions which were extracted by the Princals analysis were 0.3833 and 0.1124 respectively. Fig. 4.17A shows the object scores of the respondents on the first and second dimensions, and it is clear from this graph that 21 of the respondents are clustered closely together. In order to investigate any further variation which might exist in this large cluster, various selections were made of the dependent variables shown in Fig. 4.17B and further Princals analyses were carried out using these selections.

The variable selections were based on the distal position of the variables in graph Fig. 4.17B and the meaning which they contributed to a description of the respondents, as described in Section 4.4.2. Combinations of between four and seven variables were used in a number of analyses and the clustering of the respondents examined. The trend which was detected during this statistical investigation was that the main group of 21 respondents was split into two groups in several of these analyses. The results of a representative example of one of these analyses will now be presented in order to clarify the variation which existed in the dependent variables of the respondents from Goeree. Fig 4.18 shows the results of a Princals analysis of four selected dependent variables and from this graph five groups of respondents were identified (Fig. 4.18A). The eigenvalues for dimensions one and two for this analysis were 0.6135 and 0.2792 respectively.

These groups are interpreted on the basis of the four variables and their component loadings are shown in Table 4.14. The four variables are as follows: the current fishing method of the respondents (curmet24), the respondent's rating of the exterior quality of his catch (qualx371), who takes the decisions on board the fishing boat (nowdec84) and where does the respondent auction his catch (auctn108).

Table 4.14: Values of four selected Dependent Variables from Goeree which were used in a Princals analysis to group the respondents.

Group	Respon.	curmet24	qualx371	nowdec84	Auctn108
	7	12beam	top	skip/consult	Stel+Sch
I	11	9shrimpbeam	top	skip/consult	Stel+Sch
	14	12beam	top	skip/consult	Stel+Sch
	4	matbeam	top	skipper	Sch
	2	bords+shrimpbeam	reas	skip/consult	Stel
	6	12beam	good	skip/consult	IJm+St
	15	9shrimpbeam	good	skip/consult	Stel+Sch
	19	beam+board	good	skip/consult	StelSchIIJmDenem
	21	12beam	good	skip/consult	Stel
II	23	9shrimpbeam	good	skip/consult	Stel
	25	12beam	good	skip/consult	Stel
	26	matbeam	reas	skip/consult	Stel
	28	12beam	good	skip/consult	Sch
	31	12beam	reas	skip/consult	Stel
	1	12beam	reas	skipper	Stel, Sch
	3	bords+shrimpbeam	good	skipper	Stel
	8	12beam	good	skipper	Stel+Sch
	9	beam+pair	good	skipper	Stel+Sch
	10	9shrimpbeam	good	skipper	Stel+Sch
III	13	12beam	good	skipper	Stel
	17	12beam	good	skipper	Stel
	18	12beam	good	skipper	Stel
	22	12beam	reas	skipper	Stel+Sch
	24	beam+pair	good	skipper	Stel+dealer
	30	12beam	good	skipper	Stel
	5	9shrimpbeam	good	50:50	Stel
IV	12	9shrimpbeam	good	50:50	StelSchIIJmDenem
	27	doubshlsuck	good	50:50	factory
	16	angling	top	skipper	anglers
V	20	angling	top	skipper	anglers
	29	doubshlsuck	top	skipper	dealer

Group I is similar to group II with regard to the fishing methods and the decision taking. Both groups had respondents who fished for shrimp with the smaller nine meter wide beam trawls and respondents who fished with the full size 12 meter beam trawl, which is primarily for sole and plaice. In group I one of the three respondents fished for shrimp and in group II three of the ten respondents fished for shrimp. Both group also took decisions on board in consultation with crew members. However, group I rated their catch as top quality while group II rated the exterior quality of their catch as reasonable and good. All the respondents in group I auction their catch in Scheveningen as well as the Goeree fleet's home auction centre in the new Delta outer harbour in Stellendam. Six of the ten respondents in group II auction their catch in Stellendam only.
Groups II and III represent a major part of the variability in the dependent variables of the Goeree data, as explained above. These two groups have one major difference, which is that all of the respondents in group II consult crew before taking decisions on board, while on all the boats in group III only the skipper takes the decisions. A minor difference between groups II and III is that group III is slightly more dominated by the large beam trawlers (nine out of 11) compared to group II which has proportionally slightly more shrimp beamers (three out of ten), however this may not represent a real difference in fishing practice between these two groups. Their assessment of the exterior quality of their catch and the auction centres which they use for selling their catch are more or less the same.
Group IV consists of two shrimp fishermen, who make decisions on board in consultation with the crew. Like groups II and III, the respondents in group IV assessed the exterior quality of their catch as good and land their catch in other centres as well as Stellendam.
Group V is made up of two angling skippers and a cockle fisherman. This group also rates the exterior appearance of its catch as top quality. In all three cases in group V the skipper makes the decisions on board and none use any auction centre to dispose of their catch. Respondents 4 and 27 do not cluster close to others in Fig. 4.18A. Respondent 4 fished with a 12 meter beam trawl which was adapted to cover rough, stony, ground with a mat of chains to hold it down and stop stones going into the net as well as the beam getting snagged on the bottom. Unlike those in groups I and II, the skipper takes the decisions on board and unlike group III he rated the exterior appearance of his catch as top quality. Respondent 4 lands his catch in Scheveningen. Respondent 27, unlike group IV, is a cockle fisherman, fishing with double sucking tubes which pump the cockles up into the boat and does not use auction centres to sell the cockles, as these go directly to the factory. In contrast to those in group V, he judged his catch's external appearance to be good and consults his crew when taking decisions.
Overall all respondents who judged the external appearance of their catch to be top quality were in the top left hand quarter of the graph in Fig. 4.18A. None of the respondents which plotted to the left of the Y axis (i.e.: with negative values on dimension one) sold their catch exclusively through the fish auction centre at Stellendam.

4.5.7: Combined Princals and Overals multivariate analyses of Traditionality Variables and Sustainability Variables of the Goeree data.
As in Section 4.4 on the Dingle questionnaire data, the relationship between the variables which indicate the traditionality of the respondents and the variables which indicate how sustainable the fishing practices of the respondent are, are now examined. From the Goeree questionnaire data, 11 traditionality variables (Fig. 4.19A) were selected and subjected to a Princals analysis, of which the object scores were taken as the variables for the first data set for an Overals analysis. The eigenvalues for this Princals analysis were 0.3798 and 0.1589 for the first and second dimensions respectively. A second Princals analysis was done on 12 sustainability variables (Fig. 4.19B) and the object scores from that analysis used as the second data set for the Overals analysis. The eigenvalues for this second Princals analysis were 0.3375 and 0.1781 for the first and second dimensions respectively.

The object scores for the first two Princals dimensions from both the traditionality and sustainability data sets from Goeree are listed in Table 4.15, these were categorised in the usual manner for the Overals analysis.

Table 4.15: Object scores from two Princals analyses, one on 11 Traditionality Variables and the other on 12 Sustainability Variables, for the respondents from Goeree.

Respondent	Traditionality Analysis			Sustainability Analysis
	Dimension 1	Dimension 2		Dimension 1	Dimension 2
1	-1.51	-2.14		-0.99	-0.14
2	-1.91	-1.26		1.19	1.13
3	0.01	0.07		1.76	0.21
4	0.59	0.91		-1.1	-0.83
5	0.56	1.33		0.94	-0.38
6	1.73	-1.23		-0.84	-0.05
7	0.76	-0.8		-0.49	-1.26
8	0.44	0.1		-0.6	-1.3
9	0.13	0.72		-1.08	0.95
10	1.73	-1.08		0.34	1.04
11	0.27	1.3		0.54	1.37
12	0.07	1.24		1.05	0.9
13	0.74	0.05		-0.81	-1.48
14	1.77	-2.02		-0.65	0.93
15	0.02	1.42		0.36	0.84
16	-0.63	0.85		1.63	-0.71
17	0.93	-0.18		-1.06	0.95
18	0.93	-0.18		-1.06	0.95
19	-0.25	0.57		-0.05	1.26
20	-1.58	0.14		2.19	-1.04
21	0.51	0.87		0.23	-0.6
22	-0.31	1		-0.05	-0.18
23	-1.09	0.04		1.26	1.06
24	-0.66	0.01		-0.66	-2.29
25	0.28	0.59		-1.08	0.36
26	0.13	-1.53		-0.96	0.55
27	-1.95	-1.11		1.52	-0.34
28	0.33	-1.05		-1.01	0.33
29	-1.78	-0.11		0.51	-2.13
30	-0.28	0.85		-0.22	-0.33
31	0.03	0.62		-0.83	0.24

The reason why the two sets of variables were not used directly in an Overals analysis has already been discussed. A plot of the Overals object scores for the Goeree respondents are shown in Fig. 4.20A and the component loadings for the two sets of Princals dimensions are shown in Fig.4.20B.

The eigenvalues for the Overals analysis were 0.789 and 0.528 for the first and second dimensions respectively. As in the analysis of the Dingle traditionality and sustainability variables, the two Goeree data sets for the Overals variables are the Princals object scores on dimensions one and two for the traditionality data set (Tradit1 and Tradit2) and the Princals object scores on dimensions one and two for the sustainability data set (Sustain1 and Sustain2), as is shown in Fig. 4.20B. Eight groups of respondents were identified (Fig. 4.20A) and these are now discussed on the basis of five traditionality variables (Fig. 4.19A) and six sustainability variables (Fig. 4.19B). The values of these variables for the eight groups are shown in Table 4.16.

Table 4.16: Values of five traditionality and six sustainability variables for respondents grouped on the basis of Overals analysis on the first and second dimension object scores of a Princals analysis of 11 traditionality and 12 sustainability variables from the Goeree questionaires.

Group	Resp.	Traditionality Variables						Sustainability Variables
		filear13	ancs1710	fisorg23	Toyr2413	nona48.2		curmet24	HighHP	grdson43	fisaverg	Nonoq862	tren1154
	1	1job	0	visserbond	27.5	4		12beam	1800	yes	3.4	5	stable
	2	1job	0	visserbond	27	3		bord+shribeam	300	maybe	3.6	3	stable
I	23	father	1	visserbond	3	1		9garnbeam	225	maybe	3.5	1	up
	27	1job	0	PO	33	2		doubshlsuck	800	maybe	2.5	8	down
			Mean 0.25		Mean 22.6	Mean 2.5			Mean 781		Mean 3.3	Mean 4.25
	6	father	3	federatie	35	5		12beam	2700	maybe	2.5	4	stable
	10	father	3	federatie	38	7		9shrimpbeam	300	maybe	3.7	6	up
II	14	1job	4	federatie	42	7		12beam	3600	don'tknow	3.625	5	stable
	26	1job	2	federatie	24	4		matbeam	2000	no	3	6	stable
			Mean 3		Mean 34.75	Mean 5.75			Mean 2150		Mean 3.2	Mean 5.25
	9	1job	3	visserbond	19	3		beam+pair	1800	no	3.1	2	stable
	17	1job	3	visserbond	47	3		12beam	1800	no	3.0625	4	stable
III	18	1job	3	visserbond	47	3		12beam	1800	no	3.0625	4	stable
	25	1job	3	visserbond	36	2		12beam	1800	yes	3.25	5	stable
	28	father	2	visserbond	37	4		12beam	2400	yes	3.25	5	stable
			Mean 2.8		Mean 37.2	Mean 3			Mean 1920		Mean 3.2	Mean 4
	4	father	3	federatie	40	1		matbeam	1800	yes	2.9	5	up
IV	8	father	2	federatie	47.5	3		12beam	1700	yes	3.125	1	down
	13	school	3	federatie	28	3		12beam	2000	yes	3.125	5	down
			Mean 2.7		Mean 38.5	Mean 2.3			Mean 1833		Mean 3.1	Mean 3.7
	21	father	3	visserbond	29	3		12beam	-	yes	2.8	-	down
V	24	1job	2	visserbond	29	3		beam+pair	1800	yes	1.9	5	down
			Mean 2.5		Mean 29	Mean 3			Mean 1800		Mean 2.3	Mean 5
	5	father	4	visserbond	38	2		9shrimpbeam	300	yes	2.9	4	down
	7	father	2	visserbond	38	4		12beam	2000	yes	3.4	4	down
VI	22	school+fath	3	visserbond	16	3		12beam	-	yes	2.7	4	stable
	30	family	2	visserbond	53	2		12beam	2000	don'tknow	3.125	5	down
	31	father	2	visserbond	43	3		12beam	2000	maybe	2.9	4	stable
			Mean 2.6		Mean 37.6	Mean 2.8			Mean 1575		Mean 3	Mean 4.2
	16	family	2	none	10	2		angling	370	yes	2.8	8	up
VII	20	1job	0	none	25	1		angling	300	no	2.375	8	down
	29	1job	1	Shelfishccop	23	2		doubshlsuck	150	yes	1.75	0	down
			Mean 1		Mean 19.3	Mean 1.7			Mean 273		Mean 2.3	Mean 5.3
	3	1job	3	visserbond	13.5	4		bord+shribeam	300	no	3.3	8	down
	11	father	4	visserbond	19	3		9shrimpbeam	300	no	3.1	6	stable
VIII	12	1job	4	visserbond	11	1		9shrimpbeam	300	no	3.3	5	stable
	15	father	3	visserbond	27.5	3		9shrimpbeam	300	no	3.2	5	stable
	19	father	2	visserbond	53	2		beam+board	300	no	3.3	5	stable
			Mean 3.2		Mean 24.8	Mean 2.6			Mean 300		Mean 3.2	Mean 5.8

The five traditionality variables are as follows: where or from whom did respondents learn most about fishing (filear13), how many ancestors of the respondent fished (ancs1710), membership of fishermen's organisations (fisorg23) and the number of traditional navigation methods which the respondents mentioned (nona48.2). The six sustainability variables are as follows: current fishing method of respondents (curmet24), current boat engine horse power (highhp), whether grandson will have the choice to fish as respondent fishes currently (grdson43), the average of the abundance ratings for past and present fish stocks (fisaverg), number of non-quotum fish species fished for (nonoq862) and the trend in fisheries during the 1990's (tren1154).

Group I is characterised by the lowest average number of fishing ancestors amongst the traditionality variables and the highest average rating for the fish stocks abundance in the sustainability variables. Group I is distinct from group VIII, which is plotted close to it, by its low number of fishing ancestors in the traditional variables, but group VIII also has a high average fish stocks assessment score in the sustainability variables. However, the average engine horse power in group I is medium, while that of group VIII is low, and group I thought that their grandsons might be able to fish using current methods while group VIII thought this would not be possible.
In the traditionality variables, group II had the second highest average number of ancestors who were fishermen and contrasted with group III in that all were members of the Dutch fishermen's organisation, the Federatie, while those in group III were all members of the Dutch Vissersbond. Group II had the highest average for the number of traditional navigation methods which the respondents remembered, this average for group III was the joint second highest. Both had high average years of experience in fishing, unlike group I which had a medium average for years of fishing experience. Amongst the sustainability variables, group II had the highest average engine horse power of all the groups, however group III also had a high average horse power. Likewise for the other sustainability variables, group II and III were similar, with both having high average fish abundance ratings and reporting fishing during the 1990's as having generally remained stable, just one respondent in group II said it had gone up. Group II is composed of three large beam trawlers and a shrimp fisherman, while group I had only one large trawler, two shrimp trawlers and a cockle fisherman. In Group III all were large beam trawlers. Group II was not sure about the fishing choices open to their grandsons, with one of the opinion that he would not have this choice. Group III was divided on this point, with three no's and two yes's. Group III had a high number of respondents who learned most about fishing in their first job, four out of five.
Group IV is located close to group VI on the graph in Fig. 4.20A, but differs from it in that all its respondents are members of the Federatie, while those in group VI are members of the Vissersbond. Group IV has the highest average years of fishing experience of all the groups, but group VI also has a high average for years of fishing experience. Amongst the traditionality variables both group IV and VI had learned most from their fathers and family, with school also being mentioned in the case of one respondent from each group, they both had medium average numbers of ancestors who fished and medium average numbers of traditional navigation methods which they remembered. In the sustainability variables both groups IV and VI were mainly composed of large beam trawlers, except one shrimp trawler in group VI, they had relatively high average horse powers, had medium average scores on the fish stocks abundance ratings and had medium average numbers of non-quotum fish species which they fished for. All respondents in group IV thought that their grandson would have the choice to fish like they were, but two of the five in group VI, although not negative, were not sure of this. The fishing trends during the 1990's varied between up and down in group IV, but stable and down in group VI. Group V differed from group IV in the traditionality variables in that all its respondents were in the Federatie and not the Vissersbond and their average years of fishing experience was less.
Group V, as group IV, were varied on where they learned most about fishing, had medium averages for fishing ancestors and had remembered somewhat higher numbers of traditional navigational methods. In the sustainability variables group V was similar to group IV in engine horse power, fishing methods, opinion on their grandsons' choices, numbers of non-quotum fish species they fish for and fishing trends during the 1990's. However, group V had a lower average fish stocks rating than group IV.
Group VI has already been discussed in relation to group IV.
Group VII was characterised by few ancestors who were fishermen, non membership of the two big fishermen's organisations, the lowest average number of years of fishing experience and least average numbers of traditional navigation methods mentioned. With the sustainability variables, group VII was lowest in average engine horse power and joint lowest with group V in average fish stocks abundance ratings. Group VII was made up of two angling skippers and a cockle fisherman and therefore dissimilar to group VIII closest to it on the graph. Group VII was varied on the opinion about their grandson's choices with two yes's and one no, and on fishing trends in 1990's with one up and two downs. Group VII had a high average on the number of species fished which don't have a quotum because the angling boats do not need quota, but cockles do need one.
Group VIII had the highest average number of ancestors who fished of all the groups, but a relatively low average number of years of fishing experience. The respondents in group VIII mentioned their father as well as their first job as having taught them most about fishing and moderate average numbers of traditional fishing methods. As mentioned before in relation to group I, all the respondents in group VIII were members of the Vissersbond. With regard to the sustainability variables, group VIII fished for the highest average number of species of fish for which no quota are needed. This group was also unanimous on thinking that their grandson would not have the choice to fish like they were now, unlike any other group. All of the respondents in this group fished with the smaller Eurokotters and therefore had a low average engine horse power. All, except one respondents in group VIII had found the trend of fishing during the 1990's stable, the one dissenter was of the opinion that it had gone down, and this group had a relatively high average for their fish stocks ratings.
Overall the respondents with more years of fishing experience were located to the right of the Y axis, with positive values on dimension one. The four respondents which had object scores on dimension one of less than -1 were not beam trawl fishermen and all but two of the respondents with positive objects scores on dimension one were larger beam trawlers. No large beam trawlers plotted in the top left hand quarter of the graph. As these fishing methods are engine horse power related, the horse power follows the same pattern on the graph. No other overall patterns of variation could be discerned in the graph in Fig. 4.20A.

Back to Main Page