Goherr: Fish consumption study

Question

How Baltic herring and salmon are used as human food in Baltic sea countries? Which determinants affect on people’s eating habits of these fish species?

Answer

Original questionnaire analysis results

• 13.3.2017 ----#: . These should be presented somewhere --Arja (talk) 07:39, 26 April 2017 (UTC) (type: truth; paradigms: science: comment)

Consumption amount estimates

• Model run 21.4.2017 [1] first distribution
• Model run 18.5.2017 with modelled data; with direct survey data

Do you want to use directly the survey data rather than modelled data?: No, use modelled data Yes, use survey data + Show code- Hide code # This is code Op_en7749/ on page [[Goherr: Fish consumption study#Answer]] library(OpasnetUtils) library(ggplot2) objects.latest("Op_en7749", code_name="initiate") # [[Goherr: Fish consumption study]] ovariables DecisionTableParser(opbase.data("Op_en7748", subset = "Decisions")) if(usesurvey) { objects.latest("Op_en7749", code_name="surveyjsp") # jsp ovariable directly based on survey data (N=2217) openv.setN(nrow(jsp@data)) } amount <- EvalOutput(amount) if(usesurvey) { oprint(summary(amount, marginals=c("Gender", "Country", "Fish","Ages","Foodpolicy"))) print(ggplot(amount@output, aes(x=amountResult+0.1, colour=Country))+stat_ecdf()+scale_x_log10()+facet_wrap(~ Fish)+ labs(x="Fish consumption (g /d)", y="Cumulative frequency")+theme_gray(base_size=24)) print(ggplot(amount@output, aes(x=amountResult+0.1, colour=Ages))+stat_ecdf()+scale_x_log10()+facet_grid(Country ~ Fish)+ labs(x="Fish consumption (g /d)", y="Cumulative frequency")+theme_gray(base_size=24)) print(ggplot(amount@output, aes(x=amountResult+0.1, colour=Gender))+stat_ecdf()+scale_x_log10()+facet_grid(Country ~ Fish)+ labs(x="Fish consumption (g /d)", y="Cumulative frequency")+theme_gray(base_size=24)) print(ggplot(often@output, aes(x=oftenResult+0.1, colour=Country))+stat_ecdf()+scale_x_log10()+facet_wrap(~ Fish)) print(ggplot(oftenside@output, aes(x=oftensideResult+0.1, colour=Country))+stat_ecdf()+scale_x_log10()+facet_wrap(~ Fish)) print(ggplot(much@output, aes(x=muchResult+0.1, colour=Country))+stat_ecdf()+scale_x_log10()+facet_wrap(~ Fish)) print(ggplot(muchside@output, aes(x=muchsideResult+0.1, colour=Country))+stat_ecdf()+scale_x_log10()+facet_wrap(~ Fish)) } else { oprint(summary(amount, marginals=c("Fish"))) print(ggplot(amount@output, aes(x=amountResult+0.1, colour=Fish))+stat_ecdf()+scale_x_log10()+ labs(x="Fish consumption (g /d)", y="Cumulative frequency")+theme_gray(base_size=24)) print(ggplot(often@output, aes(x=oftenResult+0.1, colour=Fish))+stat_ecdf()+scale_x_log10()) print(ggplot(oftenside@output, aes(x=oftensideResult+0.1, colour=Fish))+stat_ecdf()+scale_x_log10()) print(ggplot(much@output, aes(x=muchResult+0.1, colour=Fish))+stat_ecdf()+scale_x_log10()) print(ggplot(muchside@output, aes(x=muchsideResult+0.1, colour=Fish))+stat_ecdf()+scale_x_log10()) }

Rationale

Survey of eating habits of Baltic herring and salmon in Denmark, Estonia, Finland and Sweden has been done in September 2016 by Taloustutkimus oy. Content of the questionnaire can be accessed in Google drive. The actual data can be found from the link below (see Data).

Data

Questionnaire

Original datafile File:Goherr fish consumption.csv.

Show details

Questions in the Goherr questionnaire(-) Obs Index Full question Question Description 1 V1 Country Country 2 V2 1. What is the postal code of your residence? Postal code 3 V3 2. What is your gender? Gender 4 V4 2. How old are you? Age 5 V5 3. What is the level of your highest education? Education 6 V6 4. How would you rate the purchasing power of your household? Purchasing power 7 V7 5. Which of the following food supplements (as a separate product or as a part of some multicompound) do you eat often at least part of the year?:Fatty acids (Omega 3, DHA, EPA etc.) Omega3 8 V8 5. Which of the following food supplements (as a separate product or as a part of some multicompound) do you eat often at least part of the year?:D vitamin Vitamin D 9 V9 5. Which of the following food supplements (as a separate product or as a part of some multicompound) do you eat often at least part of the year?:Other vitamins Other vitamins 10 V10 5. Which of the following food supplements (as a separate product or as a part of some multicompound) do you eat often at least part of the year?:None of the above No vitamins 11 V11 5. Which of the following food supplements (as a separate product or as a part of some multicompound) do you eat often at least part of the year?:Don't Know Unknown vitamins 12 V12 6. Does someone in your household fish?:No No fishing 13 V13 6. Does someone in your household fish?:Yes, for recreational purposes Recreational 14 V14 6. Does someone in your household fish?:Yes, for household use Household use 15 V15 6. Does someone in your household fish?:Yes, as a professional fisherman Professional fishing 16 V16 7. Do you eat fish? Eat fish 17 V17 8. Why don't you eat fish?:Allergy Allergy 18 V18 8. Why don't you eat fish?:I don't like the taste Bad taste 19 V19 8. Why don't you eat fish?:I don't like fish bones Bones 20 V20 8. Why don't you eat fish?:Difficult to cook Cooking difficult 21 V21 8. Why don't you eat fish?Health risks Health risks fish 22 V22 8. Why don't you eat fish?:I'm not used to Not used to 23 V23 8. Why don't you eat fish?:I'm concerned about the sustainability of fish stocks Sustainability fish 24 V24 8. Why don't you eat fish?:It is ethically wrong to eat animals Ethically wrong 25 V25 8. Why don't you eat fish?:I'm a vegetarian Vegetarian 26 V26 8. Why don't you eat fish?:Don't Know Why not fish unknown 27 V27 8. Why don't you eat fish?:Other Why not fish other 28 V28 8. Why don't you eat fish?:Other open Why not fish open 29 V29 9. How often do you eat fish? How often fish 30 V30 10. Do you eat salmon or trout at least sometimes? Eat salmon 31 V31 11. Which of the following salmonids do you eat at least sometimes?:Wild Baltic salmon Baltic salmon 32 V32 11. Which of the following salmonids do you eat at least sometimes?:Wild Atlantic salmon Atlantic salmon 33 V33 11. Which of the following salmonids do you eat at least sometimes?:Norwegian cultured salmon Norwegian salmon 34 V34 11. Which of the following salmonids do you eat at least sometimes?:Rainbow trout Rainbow trout 35 V35 11. Which of the following salmonids do you eat at least sometimes?:I never know which salmon I eat Unknown salmon 36 V36 11. Which of the following salmonids do you eat at least sometimes?:Some other salmon, what? Other salmon 37 V37 11. Which of the following salmonids do you eat at least sometimes?:Other open Salmon open 38 V38 12. Where do you usually get the wild Baltic salmon that you eat?:Someone in our hosehold catches it Own catch BS 39 V39 12. Where do you usually get the wild Baltic salmon that you eat?:Buy from a grocery store Grocery store BS 40 V40 12. Where do you usually get the wild Baltic salmon that you eat?:Buy from a market place/fishmonger or similar Market place BS 41 V41 12. Where do you usually get the wild Baltic salmon that you eat?:Eat in a restaurant/cafe/canteen or similar Restaurant BS 42 V42 12. Where do you usually get the wild Baltic salmon that you eat?:Directly from a fisher Fisher BS 43 V43 12. Where do you usually get the wild Baltic salmon that you eat?:From a relative, friend, or similar Friend BS 44 V44 12. Where do you usually get the wild Baltic salmon that you eat?:Other Where BS other 45 V45 12. Where do you usually get the wild Baltic salmon that you eat?:Other open Where BS open 46 V46 13. How often do you on average eat wild Baltic salmon as an ingredient in a dish (i.e. as part of a soup, gravy, sushi, casserole etc.)? How often BS 47 V47 14. How much on average do you eat of a dish with wild Baltic salmon in it? Think of a medium sized lunch plate and the amount of the whole dish, not just the salmon. How much BS 48 V48 15. How often do you on average eat wild Baltic salmon as a side dish (i.e as a separate product such as fish balls, stake, fillets, graved, canned etc.)? How often side BS 49 V49 16. What is the average amount of wild Baltic salmon as a side dish? Think of a medium sized lunch plate. How much side BS 50 V50 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It tastes good Tastes good BS 51 V51 17. Select up to three most important reasons for you to eat wild Baltic salmon?:I caught it Self caught BS 52 V52 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It's easy to cook Easy to cook BS 53 V53 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It's quick to cook Quick to cook BS 54 V54 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It's readily available Readily available BS 55 V55 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It's healthy Healthy BS 56 V56 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It's inexpensive Inexpensive BS 57 V57 17. Select up to three most important reasons for you to eat wild Baltic salmon?:My family likes it Family likes it BS 58 V58 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It's environmentally/climate friendly choice Environmental BS 59 V59 17. Select up to three most important reasons for you to eat wild Baltic salmon?:It's a traditional dish Traditional BS 60 V60 17. Select up to three most important reasons for you to eat wild Baltic salmon?:Some other reasons, what? Why BS other 61 V61 17. Select up to three most important reasons for you to eat wild Baltic salmon?:Other open Why BS open 62 V62 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:I don't like the taste Bad taste BS 63 V63 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:I am not used to eating it / not offered at home Not used to BS 64 V64 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:I don't know how to cook it Cannot cook BS 65 V65 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:It is not easily available Not available BS 66 V66 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:I'm worried about the possible health risks caused by the harmful chemicals in it Health risks BS 67 V67 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:I doubt the quality/freshness of the products Quality issues BS 68 V68 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:I'm worried about the sustainability of the stocks Sustainability BS 69 V69 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:It doesn't belong to traditional dishes of my home area Not traditional BS 70 V70 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:It is too expensive Expensive BS 71 V71 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:Some other reason, what? Why not BS other 72 V72 18. You answered in the previous question that you don't eat wild Baltic salmon. Select up to three most important reasons for why not eating it.:Other open Why not BS open 73 V73 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Lower price Lower price BS 74 V74 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Higher price Higher price BS 75 V75 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Better availability Better availability BS 76 V76 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Availability of wider variety of ready meals / refined products Ready meals BS 77 V77 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Cooking suggestions and recipes provided within the store or in the package. Recipes BS 78 V78 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Better information (e.g. package markings or in the store) on the catch date, area, fisher and/or the refiner Package markings BS 79 V79 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Improvement of the stocks (eco-labelling) Eco labelling BS 80 V80 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Lower level of harmful chemicals in the fish Less chemicals BS 81 V81 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:National food safety authorities publish a recommendation to eat it Recommended BS 82 V82 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:National food safety authorities publish a recommendation to limit eating it Not recommended BS 83 V83 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Some other reason, what? Change BS other 84 V84 19. How would the following factors influence on your consumption of wild Baltic salmon compared to your current use? Select one for each option:Other open Change BS open 85 V85 20. Do you eat any herring (Baltic or other) at least sometimes? Eat herring 86 V86 21. Do you eat Baltic herring? This means herring caught from Baltic sea, not from Atlantic ocean Eat BH 87 V87 22. Where do you usually get the wild Baltic herring that you eat?:Someone in our hosehold catches it Own catch BH 88 V88 22. Where do you usually get the wild Baltic herring that you eat?:Buy from a grocery store Grocery store BH 89 V89 22. Where do you usually get the wild Baltic herring that you eat?:Buy from a market place/fishmonger or similar Market place BH 90 V90 22. Where do you usually get the wild Baltic herring that you eat?:Eat in a restaurant/cafe/canteen or similar Restaurant BH 91 V91 22. Where do you usually get the wild Baltic herring that you eat?:Directly from a fisher Fisher BH 92 V92 22. Where do you usually get the wild Baltic herring that you eat?:From a relative, friend, or similar Friend BH 93 V93 22. Where do you usually get the wild Baltic herring that you eat?:Other Where BH other 94 V94 22. Where do you usually get the wild Baltic herring that you eat?:Other open Where BH open 95 V95 23. How often do you on average eat Baltic herring as ingredient of a dish? (i.e as part of a dish such as in a casserole or some other food including Baltic herring)? How often BH 96 V96 24. How much on average do you eat of a dish with Baltic herring in it? Think of a medium sized lunch plate and the amount of the whole dish, not just the amount of Baltic herring How much BH 97 V97 25. How often do you on average eat whole Baltic herrings or fillets as a side dish (fried, smoked, rolls, pickled, salted or otherwise prepared)? How often side BH 98 V98 26. How much on average of Baltic herring as a side dish (think of a regular sized lunch plate)? How much side BH 99 V99 27. Select up to three most important reasons for you to eat Baltic herring?:It tastes good Tastes good BH 100 V100 27. Select up to three most important reasons for you to eat Baltic herring?:I caught it Self caught BH 101 V101 27. Select up to three most important reasons for you to eat Baltic herring?:It's easy to cook Easy to cook BH 102 V102 27. Select up to three most important reasons for you to eat Baltic herring?:It's quick to cook Quick to cook BH 103 V103 27. Select up to three most important reasons for you to eat Baltic herring?:It's readily available Readily available BH 104 V104 27. Select up to three most important reasons for you to eat Baltic herring?:It's healthy Healthy BH 105 V105 27. Select up to three most important reasons for you to eat Baltic herring?:It's inexpensive Inexpensive BH 106 V106 27. Select up to three most important reasons for you to eat Baltic herring?:My family likes it Family likes it BH 107 V107 27. Select up to three most important reasons for you to eat Baltic herring?:It's environmentally/climate friendly choice Environmental BH 108 V108 27. Select up to three most important reasons for you to eat Baltic herring?:It's a traditional dish Traditional BH 109 V109 27. Select up to three most important reasons for you to eat Baltic herring?:Some other reasons, what? Why BH other 110 V110 27. Select up to three most important reasons for you to eat Baltic herring?:Other open Why BH open 111 V111 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:I don't like the taste and/or smell of it Bad taste BH 112 V112 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:I am not used to eating it / not offered at home Not used to BH 113 V113 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:I don't know how to cook it Cannot cook BH 114 V114 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:It is difficult to cook Difficult to cook BH 115 V115 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:It is not easily available Not available BH 116 V116 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:I'm worried about the possible risks caused by the harmful chemicals in it Health risks BH 117 V117 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:It is more suitable for animal feed than human food Better for feed BH 118 V118 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:I doubt the quality/freshness of the products Quality issues BH 119 V119 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:I'm worried about the sustainability of the stocks Sustainability BH 120 V120 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:It doesn't belong to traditional dishes of my home area Not traditional BH 121 V121 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:Some other reason, what? Why not BH other 122 V122 28. You answered in the previous question that you don't eat Baltic herring. Select up to three most important reasons for why not eating it.:Other open Why not BH open 123 V123 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Lower price Lower price BH 124 V124 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Higher price Higher price BH 125 V125 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Better availability Better availability BH 126 V126 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Availability of wider variety of ready meals / processed products Ready meals BH 127 V127 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Cooking suggestions and recipes provided in the store or in the package. Recipes BH 128 V128 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Better information (e.g. package markings or in the store) on the catch date, area, fisher and/or processing company. Package markings BH 129 V129 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Improved stocks (eco-labelling) Eco labelling BH 130 V130 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Lower level of harmful chemicals in the fish Less chemicals BH 131 V131 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:National food safety authorities publish a recommendation to eat it Recommended BH 132 V132 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:National food safety authorities publish a recommendation to limit eating it Not recommended BH 133 V133 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Some other reason, what? Change BH other 134 V134 29. How would the following factors influence your consumption of Baltic herring compared to your current use? Select one option for each row.:Other open Change BH open 135 V135 30. Are you familiar with the possible eating recommendations of Baltic herring and/or wild Baltic salmon in your country? Recommendation awareness 136 V136 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It contributes to the economic wellbeing of the society Economic wellbeing BS 137 V137 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It is a natural resource for humans to utilize Natural resource BS 138 V138 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It belongs to all Baltic Sea citizens equally Belongs to all BS 139 V139 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It involves long traditions related to fishing and fish eating Long traditions BS 140 V140 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It inspires to search experiences from nature, cultural sites, fish markets or from different dishes Inspiring experiences BS 141 V141 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It is among the most renowned fish species of the Baltic Sea Renowned species BS 142 V142 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It is part of the Baltic Sea ecosystem Part of ecosystem BS 143 V143 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:I have an emotional bond to it Emotional bond BS 144 V144 31. How strongly do you agree or disagree with the following statements on Baltic salmon? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It helps me to achieve my personal aims Personal aims 145 V145 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It contributes to the economic wellbeing of the society Economic wellbeing BH 146 V146 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It is a natural resource for humans to utilize Natural resource BH 147 V147 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It belongs to all Baltic Sea citizens equally Belongs to all BH 148 V148 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It involves long traditions related to fishing and fish eating Long traditions BH 149 V149 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It inspires to search experiences from nature, cultural sites, fish markets or from different dishes Inspiring experiences BH 150 V150 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It is among the most renowned fish species of the Baltic Sea Renowned species BH 151 V151 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It is part of the Baltic Sea ecosystem Part of ecosystem BH 152 V152 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:I have an emotional bond to it Emotional bond BH 153 V153 32. How strongly do you agree or disagree with the following statements on Baltic herring? Please indicate your position by selecting a number from -5 (strongly disagree) to 5 (strongly agree).:It helps me to achieve my personal aims Personal aims BH 154 V154 Weighting Weighting 155 V155 How old are you? Age category FI Age category FI A derived variable 156 V156 2. How old are you? Age category SWE, DK, EST Age category other A derived variable 157 V157 Row number of respondent Row A derived variable 158 V158 Age category "18-45" or ">45" Ages A derived variable 159 V159 Original question converted to a number and .n added to name Eat.fish.m A derived variable 160 V160 Original question converted to a number and .n added to name How.often.fish.n A derived variable 161 V161 Original question converted to a number and .n added to name Eat.salmon.n A derived variable 162 V162 Original question converted to a number and .n added to name Baltic.salmon.n A derived variable 163 V163 Original question converted to a number and .n added to name How.often.BS.n A derived variable 164 V164 Original question converted to a number and .n added to name How.much.BS.n A derived variable 165 V165 Original question converted to a number and .n added to name How.often.side.BS.n A derived variable 166 V166 Original question converted to a number and .n added to name How.much.side.BS.n A derived variable 167 V167 Original question converted to a number and .n added to name Better.availability.BS.n A derived variable 168 V168 Original question converted to a number and .n added to name Less.chemicals.BS.n A derived variable 169 V169 Original question converted to a number and .n added to name Eat.BH.n A derived variable 170 V170 Original question converted to a number and .n added to name How.often.BH.n A derived variable 171 V171 Original question converted to a number and .n added to name How.much.BH.n A derived variable 172 V172 Original question converted to a number and .n added to name How.often.side.BH.n A derived variable 173 V173 Original question converted to a number and .n added to name How.much.side.BH.n A derived variable 174 V174 Original question converted to a number and .n added to name Better.availability.BH.n A derived variable 175 V175 Original question converted to a number and .n added to name Less.chemicals.BH.n A derived variable

Assumptions

The following assumptions are used:

Assumptions for calculations(-)

Obs	Variable	Value	Unit	Result	Description
1	freq	1	times /a	0	Never
2	freq	2	times /a	0.5 - 0.9	less than once a year
3	freq	3	times /a	2 - 5	A few times a year
4	freq	4	times /a	12 - 36	1 - 3 times per month
5	freq	5	times /a	52	once a week
6	freq	6	times /a	104 - 208	2 - 4 times per week
7	freq	7	times /a	260 - 364	5 or more times per week
8	amdish	1	g /serving	20 - 70	1/6 plate or below (50 grams)
9	amdish	2	g /serving	70 - 130	1/3 plate (100 grams)
10	amdish	3	g /serving	120 - 180	1/2 plate (150 grams)
11	amdish	4	g /serving	170 - 230	2/3 plate (200 grams)
12	amdish	5	g /serving	220 - 280	5/6 plate (250 grams)
13	amdish	6	g /serving	270 - 400	full plate (300 grams)
14	amdish	7	g /serving	400 - 550	overly full plate (500 grams)
15	ingredient		fraction	0.1 - 0.3	Fraction of fish in the dish
16	amside	1	g /serving	20 - 70	1/6 plate or below (50 grams)
17	amside	2	g /serving	70 - 130	1/4 plate (100 grams)
18	amside	3	g /serving	120 - 180	1/2 plate (150 grams)
19	amside	4	g /serving	170 - 230	2/3 plate (200 grams)
20	amside	5	g /serving	220 - 280	5/6 plate (250 grams)
21	change	1	fraction	-1 - -0.8	Decrease it to zero
22	change	2	fraction	-0.9 - -0.5	Decrease it to less than half
23	change	3	fraction	-0.6 - -0.1	Decrease it a bit
24	change	4	fraction	0	No effect
25	change	5	fraction	0.1 - 0.6	Increase it a bit
26	change	6	fraction	0.5 - 0.9	Increase it over by half
27	change	7	fraction	0.8 - 1.3	Increase it over to double
28	change	8	fraction	-0.3 - 0.3	Don't know

Preprocessing

This code is used to preprocess the original questionnaire data from the above .csv file and to store the data as a usable variable to Opasnet base. The code stores a data.frame named survey.

• Model run 11.7.2018 [2]

+ Show code- Hide code # This code is Op_en7749/preprocess2 on page [[Goherr: Fish consumption study]] library(OpasnetUtils) objects.latest("Op_en6007", code_name = "webropol") # [[OpasnetUtils/Drafts]] webropol.convert, merge.questions ############# Data preprocessing # Get the data either from Opasnet or your own hard drive. #survey1 original file: N:/Ymal/Projects/Goherr/WP5/Goherr_fish_consumption.csv survey1 <- opasnet.csv( "5/57/Goherr_fish_consumption.csv", wiki = "opasnet_en", sep = ";", fill = TRUE, quote = "\"" ) #survey1 <- re#ad.csv(file = "N:/Ymal/Projects/Goherr/WP5/Goherr_fish_consumption.csv", # header=FALSE, sep=";", fill = TRUE, quote="\"") # Data file is converted to data.frame using levels at row 2121. survey1 <- webropol.convert(survey1, 2121, textmark = ":Other open") # Delete rows that are clearly overestimations survey1 <- survey1[-c(636, 1465, 1865, 1876, 2062, 2088), ] # Take the relevant columnames from the table on the page. colnames(survey1) <- gsub( " ", ".", opbase.data( "Op_en7749", subset = "Questions in the Goherr questionnaire")$Result[1:ncol(survey1)] ) survey1$Row <- 1:nrow(survey1) survey1$Weighting <- as.double(gsub(",",".", survey1$Weighting)) survey1$Ages <- factor( ifelse(as.numeric(as.character(survey1$Age)) < 46, "18-45",">45"), levels = c("18-45", ">45"), ordered = TRUE ) # Anonymize data survey1$Postal.code <- NA survey1$Age <- NA # webropol.convert should put these in the right order but doesn't. So do it manually. freqlist <- c( "less than once a year", "A few times a year", "1 - 3 times per month", "once a week", "2 - 4 times per week", "5 or more times per week" ) amlist <- c( "1/6 plate or below (50 grams)", "1/3 plate (100 grams)", "1/2 plate (150 grams)", "2/3 plate (200 grams)", "5/6 plate (250 grams)", "full plate (300 grams)", "overly full plate (500 grams)" # "Not able to estimate" ) sidel <- c( "1/6 plate or below (50 grams)", "1/4 plate (100 grams)", "1/2 plate (150 grams)", "2/3 plate (200 grams)", "5/6 plate (250 grams)" # "Not able to estimate" ) fishamounts <- c(29,46:49,95:98) colnames(survey1)[fishamounts] #[1] "How.often.fish" "How.often.BS" "How.much.BS" "How.often.side.BS" #[5] "How.much.side.BS" "How.often.BH" "How.much.BH" "How.often.side.BH" #[9] "How.much.side.BH" ansl <- list( freqlist, c("Never", freqlist), amlist, c("Never", freqlist), sidel, c("Never", freqlist), amlist, c("Never", freqlist), sidel ) for (i in 1:length(fishamounts)) { survey1[[fishamounts[i]]] <- factor(survey1[[fishamounts[i]]], levels = ansl[[i]], ordered = TRUE) } oprint(head(survey1)) agel <- as.character(unique(survey1$Ages)) countryl <- sort(as.character(unique(survey1$Country))) genderl <- sort(as.character(unique(survey1$Gender))) fisl <- c("Salmon", "Herring") # Interesting fish eating questions surv <- c(1,3,158,16,29,30,31,46:49,75,80,86,95:98, 125, 130) colnames(survey1[surv]) # Shown without .n #[1] "Country" "Gender" #[3] "Ages" "Fish eating" #[5] "How often eat fish" "Salmon eating" #[7] "Baltic salmon" "How often Baltic salmon" #[9] "How much Baltic salmon" "How often side Baltic salmon" #[11] "How much side Baltic salmon" "Better availability BS" #[13] "Less chemicals BS" "Eat Baltic herring" #[15] "How often Baltic herring" "How much Baltic herring" #[17] "How often side Baltic herring" "How much side Baltic herring" #[19] "Better availability BH" "Less chemicals BH" # For estimating distributions, we should #1 remove people with Fish eating = No (142) #2 merge Eat Baltic herring = I don't know with No (How often BH = NA always) #3 merge Baltic salmon = NA with No (because they usually have answered BH questions) oprint(table(is.na(rowSums(sapply(survey1[surv[4:20]], as.numeric))))) # BUT: there are so many missing values, that we just model BH and BS separately now. # Take the key questions and treat them as numeric temp <- as.data.frame(lapply( survey1[c(16,29,30,31,46:49,75,80,86,95:98, 125, 130)], FUN = function(x) as.integer(x) )) # Coerce to integers colnames(temp) <- paste(colnames(temp),"n",sep=".") survey1 <- cbind(survey1,temp) survey1[is.na(survey1$Eat.Baltic.herring.n) | survey1$Eat.Baltic.herring.n==3 , 14] <- 1 # I don't know --> No # Logical variable for respondents that have eaten fish, Baltic salmon, and Baltic herring survey1$Eatfish <- survey1$Eat.fish == "Yes" survey1$Eatsalm <- survey1$Baltic.salmon=="Yes" & !is.na(rowSums(survey1[162:166])) survey1$Eatherr <- survey1$Eat.BH=="Yes" & !is.na(rowSums(survey1[169:173])) oprint(table(survey1[c("Eatsalm", "Eatherr", "Eatfish")], useNA = "ifany")) # Oletetaan, että covarianssimatriisi on vakio kaikille maille ja sukupuolille yms # mutta keskiarvo on spesifi näille ja kysymykselle. agel countryl genderl fisl objects.store(survey1, surv, agel, countryl, genderl, fisl) cat("Data.frame survey1 and vectors agel, countryl, genderl and fisl were stored.\n")

Analyses

• Sketches about modelling determinants of eating (spring 2018) [3]

Figures, tables and stat analyses for the first manuscript

• Model run 15.6.2018 [4]
• Model run 16.6.2018 [5]
• Model run 20.6.2018 [6]
• Model run 11.7.2018 [7]
• Model run 9.8.2018 [8]
• Model run 12.8.2018 [9]
• Model run 12.9.2018 with country-specific regression analyses, open-ended asnwers and fig about eating any herring [10]

+ Show code- Hide code # This is code Op_en7749/ on page [[Goherr: Fish consumption study]] library(OpasnetUtils) library(ggplot2) library(reshape2) library(MASS) #library(car) #library(vegan) BS <- 24 groups <- function(o) { o$Group <- paste(o$Gender, o$Ages) o$Group <- factor(o$Group, levels = c( "Female 18-45", "Male 18-45", "Female >45", "Male >45" )) return(o) } #' @title reasons produces a ggplot object for a graph that shows fractions of reasons mentioned in the data. #' @param dat a data.frame containing survey answers for the target subgroup. #' @param title an atomic string with the name to be shown as the title of graph reasons <- function( dat, title ) { weight <- Ovariable("weight",data=data.frame( Row=dat$Row, Result=dat$Weight )) levels(dat$Country) <- paste0( levels(dat$Country), " (n=", aggregate(dat$Country, dat["Country"], length)$x, ")" ) tmp <- melt( dat[setdiff(colnames(dat),"Weighting")], id.vars = c("Country","Row"), variable.name ="Reason", value.name="Result" ) tmp$Result <- (tmp$Result %in% c("Yes","1"))*1 tmp <- weight * (Ovariable("tmp",data=tmp)) cat(title, ": number of individual answers.\n") oprint(oapply(tmp, c("tmpResult","Country"),length)@output) tmp <- (oapply(tmp, c("Country","Reason"), sum)/oapply(tmp, "Country",sum)) popu <- oapply(tmp, "Reason",sum)@output popu <- popu$Reason[order(popu$Result)] tmp$Reason <- factor(tmp$Reason, levels=popu) levels(tmp$Reason) <- gsub("( BH| BS)", "", gsub("\\.", " ", levels(tmp$Reason))) cat(title, ": fractions shown on graph.\n") oprint(tmp@output) tmp <- ggplot(tmp@output, aes(x=Reason, y=Result,colour=Country, group=Country))+ geom_point(shape=21, size=6, fill="Grey", stroke=2)+ geom_line()+ coord_flip()+ theme_gray(base_size=BS)+ scale_y_continuous(labels=scales::percent_format())+ labs( title=title, x="Answer", y="Fraction of population") return(tmp) } impacts <- function(dat,title) { tmp <- melt( dat, id.vars = "Country", variable.name ="Reason", value.name="Response" ) tmp$Response <- factor( tmp$Response, levels=c( "I don't know", "Increase it over to double", "Increase it over by half", "Increase it a bit", "No effect", "Decrease it a bit", "Decrease it to less than half", "Decrease it to zero" ), ordered=TRUE ) levels(tmp$Response)[levels(tmp$Response) %in% c( "Decrease it to zero", "Decrease it to less than half", "Decrease it a bit" )] <- "Decrease" levels(tmp$Response)[levels(tmp$Response) %in% c( "Increase it a bit", "Increase it over by half", "Increase it over to double" )] <- "Increase" popu <- aggregate(as.numeric(tmp$Response), tmp["Reason"],sum) popu <- popu$Reason[order(popu$x)] tmp$Reason <- factor(tmp$Reason, levels=popu) levels(tmp$Reason) <- gsub("( BH| BS)", "", gsub("\\.", " ", levels(tmp$Reason))) tmp$Result <- 1 tmp <- EvalOutput(Ovariable("tmp",data=tmp)) tmp <- (oapply(tmp, c("Country","Reason","Response"), sum)/ oapply(tmp, c("Country","Reason"),sum)) tmp <- ggplot(tmp@output, aes(x=Reason, weight=Result, fill=Response))+geom_bar()+ coord_flip()+facet_grid(.~Country)+ theme_gray(base_size=BS)+ theme(legend.position = "bottom")+ scale_fill_manual(values=c("Grey","Green","Yellow","Red"))+ scale_y_continuous(breaks=c(0.5,1),labels=scales::percent_format())+ labs( title=title, x="Cause", y="Fraction of population" ) return(tmp) } ################## Get data objects.latest("Op_en7749", "preprocess2") # [[Goherr: Fish consumption study]]: survey1, surv objects.latest("Op_en7749", "nonsamplejsp") # [[Goherr: Fish consumption study]]: jsp every respondent from data exactly once. objects.latest("Op_en7749", "initiate") # [[Goherr: Fish consumption study]]: amount etc. openv.setN(50) survey1 <- groups(survey1) effinfo <- 0 # No policies implemented. effrecomm <- 0 amount <- EvalOutput(amount) #result(amount)[result(amount)==0] <- 0.1 Not needed if not log-transformed. #### General fish eating su1 <- EvalOutput(Ovariable( "su1", data = data.frame( survey1[c("Country","Gender","Education","Purchasing.power")], Result=1 ) )) #table 2: statistics of the survey population in each country (n, female %, education, purchasing power) tmp <- round(t(data.frame( n = tapply(rep(1,nrow(survey1)), survey1$Country,sum), Females = tapply(survey1$Gender=="Female", survey1$Country,mean)*100, Old = tapply(survey1$Ages==">45", survey1$Country,mean)*100, PurcVlow = tapply(survey1$Purchasing.power=="Very low", survey1$Country,mean)*100, PurcLow = tapply(survey1$Purchasing.power=="Low", survey1$Country,mean)*100, PurcSuf = tapply(survey1$Purchasing.power=="Sufficient", survey1$Country,mean)*100, PurcGood = tapply(survey1$Purchasing.power=="Good", survey1$Country,mean)*100, PurcVgood = tapply(survey1$Purchasing.power=="Very good", survey1$Country,mean)*100, PurcExc = tapply(survey1$Purchasing.power=="Excellent", survey1$Country,mean)*100, EducPri = tapply(survey1$Education=="Primary education", survey1$Country,mean)*100, EducSec = tapply(survey1$Education=="Secondary education (gymnasium, vocational school or similar)", survey1$Country,mean)*100, EducCol = tapply(survey1$Education=="Lower level college education or similar", survey1$Country,mean)*100, EducHig = tapply(survey1$Education=="Higher level college education or similar", survey1$Country,mean)*100 ))) rownames(tmp) <- c( "Number of respondents", "Females (%)", ">45 years (%)", "Purchasing power: Very low (%)", "Low (%)", "Sufficient (%)", "Good (%)", "Very good (%)", "Excellent (%)", "Education: Primary education (%)", "Secondary education (%)", "Lower level college (%)", "Higher level college (%)" ) oprint(tmp, digits=0) cat("Weighted fraction of fish eaters\n") tmp <- aggregate(survey1$Weighting, survey1[c("Eat.fish","Country")],FUN=sum) oprint(aggregate(tmp$x, tmp["Country"],FUN=function(x) x[2]/sum(x))) ggplot(survey1, aes(x=Eat.fish, weight=Weighting, group=Country), stat="count")+ geom_bar(aes(y=..prop.., fill=Country), position="dodge")+ theme_gray(base_size=BS)+ scale_y_continuous(labels=scales::percent_format())+ labs(title="Fraction of fish eaters within population") ############## Answers to open questions cat("Number of open-ended answers per question and country\n") tmp <- survey1[c(1,3,27:28,36:37,44:45,60:61,71:72,83:84,93:94,109:110,121:122,133:134,154,157:158)] tmp <- reshape( tmp, varying=list( Condition = colnames(tmp)[grepl("other",tolower(colnames(tmp)))], Text = colnames(tmp)[grepl("open",colnames(tmp))] ), times = colnames(tmp)[grepl("other",tolower(colnames(tmp)))], direction="long" ) tmp <- tmp[tmp$Why.not.fish.open!="" , ] oprint(table(tmp$time,tmp$Country)) oprint(tmp) #### Baltic salmon reasons( survey1[survey1$Eat.fish=="No"& !is.na(survey1$Eat.fish),c(1,17:26,154,157)], # 1 Country, 8 Why don't you eat fish "Reasons not to eat among fish non-consumers" ) reasons( survey1[survey1$Eat.salmon=="Yes"& !is.na(survey1$Eat.salmon),c(1,31:36,154,157)], # 1 Country, 11 Which salmon species "Species used among salmon consumers" ) reasons( survey1[survey1$Baltic.salmon=="Yes"& !is.na(survey1$Baltic.salmon),c(1,50:59,154,157)], # 1 Country, 17 Why Baltic salmon "Reasons to eat among Baltic salmon consumers" ) reasons( survey1[survey1$Baltic.salmon=="No"& !is.na(survey1$Baltic.salmon),c(1,62:70,154,157)], # 1 Country, 18 Why not Baltic salmon "Reasons not to eat among Baltic salmon non-consumers" ) impacts( survey1[survey1$Baltic.salmon=="Yes"& !is.na(survey1$Baltic.salmon),c(1,73:82)], "Effect on Baltic salmon consumption" ) ####### Any herring ggplot(survey1[survey1$Eat.fish=="Yes",], aes(x=Country, weight=Weighting, fill=Eat.herring))+geom_bar(position="fill")+ coord_flip()+#facet_grid(.~Country)+ theme_gray(base_size=BS)+ theme(legend.position = "bottom")+ scale_y_continuous(labels=scales::percent_format())+ # scale_fill_manual(values=c("Grey","Green","Yellow","Red"))+ labs( title="Any herring consumption", y="Fraction of fish consumers" ) ####### Baltic herring ggplot(survey1[survey1$Eat.fish=="Yes",], aes(x=Country, weight=Weighting, fill=Eat.BH))+geom_bar(position="fill")+ coord_flip()+#facet_grid(.~Country)+ theme_gray(base_size=BS)+ theme(legend.position = "bottom")+ scale_y_continuous(labels=scales::percent_format())+ # scale_fill_manual(values=c("Grey","Green","Yellow","Red"))+ labs( title="Baltic herring consumption", y="Fraction of fish consumers" ) ggplot(survey1[survey1$Eat.herring=="Yes",], aes(x=Country, weight=Weighting, fill=Eat.BH))+geom_bar(position="fill")+ coord_flip()+#facet_grid(.~Country)+ theme_gray(base_size=BS)+ theme(legend.position = "bottom")+ scale_y_continuous(labels=scales::percent_format())+ # scale_fill_manual(values=c("Grey","Green","Yellow","Red"))+ labs( title="Baltic herring consumption", y="Fraction of herring consumers" ) levels(survey1$Recommendation.awareness) <- c("Not aware","Aware","Know content") ggplot(survey1, aes(x=Country, weight=Weighting, fill=Recommendation.awareness))+geom_bar(position="fill")+ coord_flip()+#facet_grid(.~Country)+ theme_gray(base_size=BS)+ theme(legend.position = "bottom")+ scale_y_continuous(labels=scales::percent_format())+ # scale_fill_manual(values=c("Grey","Green","Yellow","Red"))+ labs(title="Awareness of food recommendations about Baltic fish") reasons( survey1[survey1$Eat.BH=="Yes"& !is.na(survey1$Eat.BH),c(1,99:108,154,157)], # 1 Country, 27 why Baltic herring "Reasons to eat among Baltic herring consumers" ) reasons( survey1[survey1$Eat.BH=="No"& !is.na(survey1$Eat.BH),c(1,111:120,154,157)], # 1 Country, 28 Why not Baltic herring "Reasons not to eat among Baltic herring non-consumers" ) impacts( survey1[survey1$Eat.BH=="Yes"& !is.na(survey1$Eat.BH),c(1,123:132)], # 1 Country, 29 Influence of Baltic herring policies title="Effect on Baltic herring consumption" ) ######### Amounts estimated for each respondent amount <- groups(amount) #tmp <- amount*info #tmp <- tmp[ # tmp$Info.improvements=="BAU" & # tmp$Recomm.herring=="BAU" & # tmp$Recomm.salmon=="BAU" , # ] tmp <- cbind( oapply(amount, c("Country","Group","Fish"), FUN=mean)@output, SD = oapply(amount, c("Country","Gender","Ages","Fish"), FUN=sd)$amountResult ) ggplot(tmp, aes(x=Group, weight=amountResult, fill=Fish))+ geom_bar()+facet_grid(Country~Fish)+ theme_gray(base_size=BS)+ theme(axis.text.x = element_text(angle = -90))+ labs( title="Fish consumption in subgroups", y="Average consumption (g/day)") tmp$Result <- paste0(sprintf("%.1f",tmp$amountResult), " (",sprintf("%.1f", tmp$SD,1),")") tmp <- reshape(tmp, v.names="Result", timevar="Country", idvar=c("Fish","Group"),drop=c("Gender","Ages","amountResult","SD"), direction="wide") colnames(tmp) <- gsub("Result\\.", "", colnames(tmp)) cat("Average fish consumption in subgroups, mean (sd)\n") oprint(tmp) result(amount)[result(amount)==0] <- 0.1 ggplot(amount@output, aes(x=amountResult, colour=Group))+stat_ecdf()+scale_x_log10() ############################ #### Statistical analyses dat <- merge(survey1,amount@output) # These should be corrected to preprocess2. NOT ordered. dat$Country <- factor(dat$Country, ordered=FALSE) dat$Ages <- factor(dat$Ages, ordered=FALSE) dat$Gender <- factor(dat$Gender, ordered=FALSE) ggplot(dat, aes(x=amountResult, weight=Weighting, color=Country))+stat_ecdf()+facet_wrap(~Fish)+ scale_x_log10() gro <- unique(dat$Group) out <- list(data.frame(), data.frame()) for(i in unique(dat$Fish)) { for(j in unique(dat$Country)){ for(k in unique(dat$Iter)) { res <- kruskal.test(amountResult ~ Group, data=dat[ dat$Iter==k & dat$Country==j & dat$Fish==i,]) out[[1]] <- rbind(out[[1]], data.frame( test="Kruskal-Wallis", Fish=i, Country=j, Iter=k, p.value=res[[3]]) ) for(l in 1:(length(gro)-1)) { for(m in (l+1):length(gro)) { res2 <- wilcox.test( x=dat$amountResult[dat$Iter==k & dat$Country==j & dat$Fish==i & dat$Group==gro[l]], y=dat$amountResult[dat$Iter==k & dat$Country==j & dat$Fish==i & dat$Group==gro[m]], conf.int = FALSE) out[[2]] <- rbind(out[[2]], data.frame( Fish=i, Country=j, Iter=k, Pair1=gro[l], Pair2=gro[m], p.value = res2$p.value )) } } } } } ggplot(out[[1]], aes(x=paste(Fish, Country), y = p.value, colour = Country))+geom_jitter()+ geom_hline(yintercept=0.05)+coord_flip()+theme_gray(base_size=BS)+ labs(title="Kruskal-Wallis non-parametric test across all groups") ggplot(out[[2]], aes(x=paste(Fish, Country), y=p.value, colour=Pair1, shape=Pair2))+geom_jitter()+ geom_hline(yintercept=0.05)+coord_flip()+theme_gray(base_size=BS)+ labs(title="Mann-Whitney U test across all pairs of groups") cat("Kurskal-Wallis non-parametric test for 50 iterations of amount\n") oprint(aggregate(out[[1]]["p.value"], out[[1]][c("Country","Fish")], mean)) cat("Mann-Whitney U non-parametric test for 50 iterations of amount\n") oprint(aggregate(out[[2]]["p.value"], out[[2]][c("Pair1","Pair2","Country","Fish")], mean)) ###################### Logistic regression cat("###################### What explains whether people eat fish at all?\n") fit <- glm(Eat.fish ~ Ages + Gender + Country + Education + Purchasing.power, family=binomial(),data=dat[dat$Iter==1,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) for(i in unique(dat$Country)) { cat("\n#### Country-specific regression analysis:", i, "\n") fit <- glm(Eat.fish ~ Ages + Gender + Education + Purchasing.power, family=binomial(), data=dat[dat$Iter==1 & dat$Country==i,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) } cat("###################### What explains whether people eat any herring?\n") fit <- glm(Eat.herring ~ Ages + Gender + Country + Education + Purchasing.power, family=binomial(),data=dat[dat$Iter==1,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) for(i in unique(dat$Country)) { cat("\n#### Country-specific regression analysis:", i, "\n") fit <- glm(Eat.herring ~ Ages + Gender + Education + Purchasing.power, family=binomial(), data=dat[dat$Iter==1 & dat$Country==i,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) } cat("###################### What explains whether people eat Baltic herring?\n") fit <- glm(Eat.BH ~ Ages + Gender + Country + Education + Purchasing.power, family=binomial(),data=dat[dat$Iter==1,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) for(i in unique(dat$Country)) { cat("\n#### Country-specific regression analysis:", i, "\n") fit <- glm(Eat.BH ~ Ages + Gender + Education + Purchasing.power, family=binomial(), data=dat[dat$Iter==1 & dat$Country==i,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) } cat("###################### What explains whether people eat any salmon?\n") fit <- glm(Eat.salmon ~ Ages + Gender + Country + Education + Purchasing.power, family=binomial(),data=dat[dat$Iter==1,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) for(i in unique(dat$Country)) { cat("\n#### Country-specific regression analysis:", i, "\n") fit <- glm(Eat.salmon ~ Ages + Gender + Education + Purchasing.power, family=binomial(), data=dat[dat$Iter==1 & dat$Country==i,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) } cat("###################### What explains whether people eat Baltic salmon?\n") fit <- glm(Baltic.salmon ~ Ages + Gender + Country + Education + Purchasing.power, family=binomial(),data=dat[dat$Iter==1,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) for(i in unique(dat$Country)) { cat("\n#### Country-specific regression analysis:", i, "\n") fit <- glm(Baltic.salmon ~ Ages + Gender + Education + Purchasing.power, family=binomial(), data=dat[dat$Iter==1 & dat$Country==i,]) oprint(summary(fit)) step <- stepAIC(fit, direction = "both") oprint(step$anova) }

Descriptive statistics

Correlation matrix of all questions in the survey (answers converted to numbers).

Model must contain predictors such as country, gender, age etc. Maybe we should first study what determinants are important? Model must also contain determinants that would increase or decrease fish consumption. This should be conditional on the current consumption. How? Maybe we should look at principal coordinates analysis with all questions to see how they behave.

Also look at correlation table to see clusters.

Some obvious results:

• If reports no fish eating, many subsequent answers are NA.
• No vitamins correlates negatively with vitamin intake.
• Unknown salmon correlates negatively with the types of salmon eaten.
• Different age categories correlate with each other.

However, there are also meaningful negative correlations:

• Country vs allergy
• Country vs Norwegian salmon and Rainbow trout
• Country vs not traditional.
• Country vs recommendation awareness
• Allergy vs economic wellbeing
• Baltic salmon use (4 questions) vs Don't like taste and Not used to
• All questions between Easy to cook ... Traditional dish

Meaningful positive correlations:

• All questions between Baltic salmon ... Rainbow trout
• How often Baltic salmon/herring/side salmon/side herring
• How much Baltic salmon/herring/side salmon/side herring
• Better availability ... Recommendation
• All questions between Economic wellbeing...Personal aims
• Omega3, Vitamin D, and Other vitamins

Model runs

• Model run 13.3.2017
• Model run 21.4.2017 [11] old code from Answer merged to this code and debugged

+ Show code- Hide code # This is code Op_en7749/ on page [[Goherr: Fish consumption study]] library(OpasnetUtils) library(ggplot2) library(reshape2) library(car) library(vegan) objects.latest("Op_en7749", "preprocess2") # [[Goherr: Fish consumption study]]: survey, surv ############################### From a previous code on Answer for(i in c(5:6, 16, 29:30, 46:49, 85:86, 95:98, 135)) { temp <- survey[!is.na(survey[[i]]),] p <- ggplot(temp, aes(x = temp[[i]])) + geom_bar() + theme_gray(base_size = 18) + theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.4)) + labs(title = colnames(temp[i])) + xlab("") + facet_wrap(~ Country) print(p) } temp <- melt(survey, measure.vars = 145:153, variable.name = "Changing.factor", value.name = "Impact") levs <- c("-5 strongly disagree", "-4", "-3", "-2", "-1", "0 Neutral", "1", "2", "3", "4", "5 strongly agree", "I don't know") temp$Impact <- factor(temp$Impact, levels = levs, labels = levs, ordered = TRUE) ggplot(temp, aes(x = Impact, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Changing.factor)+ theme_gray(base_size = 24) + theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.4)) surveytemp <- subset(survey, survey[[31]] == "Yes") temp <- melt(surveytemp, measure.vars = 38:43, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + labs(title = "Baltic salmon sources") temp <- melt(surveytemp, measure.vars = 50:59, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + labs(title = "Reasons to eat Baltic salmon") temp <- melt(surveytemp, measure.vars = 73:82, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.4)) + labs(title = "Baltic salmon actions") surveytemp <- subset(survey, survey[[31]] == "No") temp <- melt(surveytemp, measure.vars = 62:70, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + labs(title = "Reasons not to eat Baltic salmon") surveytemp <- subset(survey, survey[[86]] == "Yes") temp <- melt(surveytemp, measure.vars = 87:92, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + labs(title = "Baltic herring sources") temp <- melt(surveytemp, measure.vars = 99:108, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + labs(title = "Reasons to eat Baltic herring") temp <- melt(surveytemp, measure.vars = 123:132, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.4)) + labs(title = "Baltic herring actions") surveytemp <- subset(survey, survey[[86]] == "No") temp <- melt(surveytemp, measure.vars = 111:120, variable.name = "Variable", value.name = "Value") ggplot(temp, aes(x = Value, fill = Country)) + geom_bar(position = "dodge") + facet_wrap(~ Variable)+ theme_gray(base_size = 24) + labs(title = "Reasons not to eat Baltic herring") #####################################3 temp <- sapply(survey, as.numeric) # Can be done for surv to get a smaller matrix survey_correlations <- (cor(temp, method="spearman", use="pairwise.complete.obs")) temp <- colnames(survey_correlations) melted_correlations <- melt(survey_correlations) melted_correlations$Var1 <- factor(melted_correlations$Var1, levels=temp) melted_correlations$Var2 <- factor(melted_correlations$Var2, levels=temp) melted_correlations$value <- ifelse(melted_correlations$value >= 0.99,NA,melted_correlations$value) ggplot(melted_correlations, aes(x = Var1, y = Var2, fill = value, label= round(value, 2)))+ geom_raster()+ theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.4))+ scale_fill_gradient2(low = "#480610", mid = "#FFFFFF", high = "#06480F", midpoint = 0, space = "Lab", guide = "colourbar") ####################### Descriptive statistics oprint(cor(surv, use = "pairwise.complete.obs")) # --> Baltic salmon and herring eating are correlated, so they should be estimated together oprint(table(surv[c(12,7,4)], useNA = "ifany")) oprint(table(surv[c(13,12)], useNA = "ifany")) ############################# Plot original data # Eating frequencies of fish and Baltic salmon and herring with random noise, all pl <- surv[surv$Eatfish,c(5,8,10,13,15)] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ggplot(data.frame( X = rep(1,5), Y = 5:1, legend = c("All", "Finland", "Sweden", "Denmark", "Estonia") ), aes(x=X, y=Y, label=legend))+ geom_text()+ labs(title="Fish eating questions with random noise") ## Fish eating questions with some random noise, all pl <- surv[surv$Eatfish,c(5,6,7,12)] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ## Fish eating questions with some random noise, FI pl <- surv[surv$Country == 1 & surv$Eatfish,c(5,6,7,12)] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ## Fish eating questions with some random noise, SWE pl <- surv[surv$Country == 2 & surv$Eatfish,c(5,6,7,12)] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ## Fish eating questions with some random noise, Dk pl <- surv[surv$Country == 3 & surv$Eatfish,c(5,6,7,12)] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ## Fish eating questions with some random noise, EST pl <- surv[surv$Country == 4 & surv$Eatfish,c(5,6,7,12)] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ## Baltic herring questions with some random noise pl <- surv[surv$Eatherr,13:16] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ## Baltic salmon questions with some random noise pl <- surv[surv$Eatsalm,8:11] scatterplotMatrix( data.matrix(pl) + runif(nrow(pl)*ncol(pl), -0.5, 0.5) ) ##################### CORRELATION MATRIX temp <- sapply(survey, as.numeric) # Can be done for surv to get a smaller matrix survey_correlations <- (cor(temp, method="spearman", use="pairwise.complete.obs")) temp <- colnames(survey_correlations) melted_correlations <- melt(survey_correlations) melted_correlations$Var1 <- factor(melted_correlations$Var1, levels=temp) melted_correlations$Var2 <- factor(melted_correlations$Var2, levels=temp) melted_correlations$value <- ifelse(melted_correlations$value >= 0.99,NA,melted_correlations$value) ggplot(melted_correlations, aes(x = Var1, y = Var2, fill = value, label= round(value, 2)))+ geom_raster()+ theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.4))+ scale_fill_gradient2(low = "#480610", mid = "#FFFFFF", high = "#06480F", midpoint = 0, space = "Lab", guide = "colourbar") ############################### PRINCIPAL COORDINATE ANALYSIS (PCoA) #tämä osa valmistaa sen datan. hypocols1 <- c(46:49,95:98) answ <- sapply(survey[hypocols1], FUN=as.numeric) answ <- as.matrix(answ[!is.na(rowSums(answ)),]) pcoa_caps <- capscale(t(answ) ~ 1, distance="euclidean") ##PCoA done ## Kuva koko hypoteeseista colstr <- c("palevioletred1","royalblue1","seagreen1","violet","khaki2","skyblue", "orange") #hypo_sizes <- (5 - colMeans(answ)) #leg_sizes <- c(4, 3, 2, 1, 0.01) #pdf(file="pcoa_plot.pdf", height=6, width=7.5) plot(pcoa_caps, display = c("sp", "wa"), type="n")#, xlim=c(-6,4.5)) ## PCoA biplot, full scale points(pcoa_caps, display= c("sp"), col="gray40") # adding the people points points(pcoa_caps, display= c("wa"), pch=19)#, cex=hypo_sizes, col=trait.cols) text(pcoa_caps, display=c("wa"), srt=25, cex=0.5) #legend(x=-6, y=3.8, levels(traits), fill=colstr, bty="n", cex=1) #legend(-6, -2, legend=c("Very likely", "Moderately likely", # "No opinion", "Moderately unlikely", "Very unlikely"), # pch=21, pt.cex = leg_sizes, bty="n", cex=1) #dev.off()

Bayes model

• Model run 3.3.2017. All variables assumed independent. [12]
• Model run 3.3.2017. p has more dimensions. [13]
• Model run 25.3.2017. Several model versions: strange binomial+multivarnormal, binomial, fractalised multivarnormal [14]
• Model run 27.3.2017 [15]
• Other models except multivariate normal were archived and removed from active code 29.3.2017.
• Model run 29.3.2017 with raw data graphs [16]
• Model run 29.3.2017 with salmon and herring ovariables stored [17]
• Model run 13.4.2017 with first version of coordinate matrix and principal coordinate analysis [18]
• Model run 20.4.2017 [19] code works but needs a safety check against outliers
• Model run 21.4.2017 [20] some model results plotted
• Model run 21.4.2017 [21] ovariables produced by the model stored.
• Model run 18.5.2017 [22] small updates
• 13.2.2018 old model run but with new Opasnet [23]

+ Show code- Hide code # This is code Op_en7749/bayes on page [[Goherr: Fish consumption study]] library(OpasnetUtils) library(ggplot2) library(reshape2) library(rjags) library(car) library(vegan) library(MASS) #library(gridExtra) # Error: package ‘gridExtra’ was built before R 3.0.0: please re-install it # Fish intake in humans # Data from data.frame survey from page [[Goherr: Fish consumption study]] objects.latest("Op_en7749", "preprocess2") # [[Goherr: Fish consumption study]]: survey, surv, ... cat("Version with multivariate normal.\n") # Development needs: ## Correlation between salmon.often and herring.often needs to be estimated. ## Gender, country and age-spesific values should be estimated. mod <- textConnection(" model{ for(i in 1:S) { survs[i,1:4] ~ dmnorm(mus[], Omegas[,]) } for(j in 1:H) { survh[j,1:4] ~ dmnorm(muh[], Omegah[,]) } mus[1:4] ~ dmnorm(mu0[1:4], S2[1:4,1:4]) Omegas[1:4,1:4] ~ dwish(S3[1:4,1:4],S) anss.pred ~ dmnorm(mus[], Omegas[,]) muh[1:4] ~ dmnorm(mu0[1:4], S2[1:4,1:4]) Omegah[1:4,1:4] ~ dwish(S3[1:4,1:4],H) ansh.pred ~ dmnorm(muh[], Omegah[,]) } ") jags <- jags.model( mod, data = list( survs = surv[surv$Eatsalm,c(8:11)], S = sum(surv$Eatsalm), survh = surv[surv$Eatherr,c(13:16)], H = sum(surv$Eatherr), mu0 = rep(2,4), S2 = diag(4)/100000, S3 = diag(4)/10000 ), n.chains = 4, n.adapt = 100 ) update(jags, 100) samps.j <- jags.samples( jags, c('mus', 'Omegas', 'anss.pred','muh','Omegah','ansh.pred'), 1000 ) js <- array( c( samps.j$mus[,,1], samps.j$Omegas[,1,,1], samps.j$Omegas[,2,,1], samps.j$Omegas[,3,,1], samps.j$Omegas[,4,,1], samps.j$anss.pred[,,1], samps.j$muh[,,1], samps.j$Omegah[,1,,1], samps.j$Omegah[,2,,1], samps.j$Omegah[,3,,1], samps.j$Omegah[,4,,1], samps.j$ansh.pred[,,1] ), dim = c(4,1000,6,2), dimnames = list( Question = 1:4, Iter = 1:1000, Parameter = c("mu","Omega1", "Omega2", "Omega3", "Omega4", "ans.pred"), Fish = c("Salmon", "Herring") ) ) # Mu for all questions about salmon scatterplotMatrix(t(js[,,1,1])) # All parameters for question 1 about salmon scatterplotMatrix(js[1,,,1]) # Same for herring scatterplotMatrix(js[1,,,2]) jsd <- melt(js) #ggplot(jsd, aes(x=value, colour=Question))+geom_density()+facet_grid(Parameter ~ Fish) #ggplot(as.data.frame(js), aes(x = anss.pred, y = Sampled))+geom_point()+stat_ellipse() coda.j <- coda.samples( jags, c('mus', 'Omegas', 'anss.pred'), 1000 ) plot(coda.j) ######## fish.param contains expected values of the distribution parameters from the model fish.param <- list( mu = apply(js[,,1,], MARGIN = c(1,3), FUN = mean), Omega = lapply( 1:2, FUN = function(x) { solve(apply(js[,,2:5,], MARGIN = c(1,3,4), FUN = mean)[,,x]) } # solve matrix: precision->covariace ) ) objects.store(fish.param) cat("List fish.param stored.\n")

Initiate ovariables

jsp taken directly from data WITHOUT salmpling.

+ Show code- Hide code # This is code Op_en7749/nonsamplejsp on page [[Goherr: Fish consumption study]] # The code produces amount esimates (jsp ovariable) directly from data rather than bayesian model or sampling. library(OpasnetUtils) jsp <- Ovariable( "jsp", dependencies = data.frame(Name="survey1", Ident="Op_en7749/preprocess2"), # [[Goherr: Food cunsumption study]] formula = function(...) { require(reshape2) sur <- survey1[c(157,1,3,158,162:178,81,82,131,132)] # Removed, not needed:16,29,30? #colnames(sur) #[1] "Row" "Country" "Gender" #[4] "Ages" "Baltic.salmon.n" "How.often.BS.n" #[7] "How.much.BS.n" "How.often.side.BS.n" "How.much.side.BS.n" #[10] "Better.availability.BS.n" "Less.chemicals.BS.n" "Eat.BH.n" #[13] "How.often.BH.n" "How.much.BH.n" "How.often.side.BH.n" #[16] "How.much.side.BH.n" "Better.availability.BH.n" "Less.chemicals.BH.n" #[19] "Eatfish" "Eatsalm" "Eatherr" #[22] "Recommended.BS" "Not.recommended.BS" "Recommended.BH" #[25] "Not.recommended.BH" # Make sure that Row is kept separate from Iter because in the sampling version they are different. # sur contained columns Eat.fish, How.often.fish, Eat.salmon. Are these needed, as all other questions are melted? No # Columns 5-7 removed, so colnames list above does not match. colnames(sur)[5] <- "Eat.BS" colnames(sur) <- gsub("\\.n","",colnames(sur)) sur[22:25] <- sapply(sur[22:25], as.numeric) sur <- melt( sur, id.vars=c(1:4), variable.name="Question", value.name="Result" ) sur$Fish <- ifelse(grepl("BH",sur$Question),"Herring","Salmon") sur$Question <- gsub("\\.BS","",sur$Question) sur$Question <- gsub("\\.BH","",sur$Question) ########## If the missing values are not adjusted, they drop out in the next stage. if(TRUE) { # The adjustments below probably should go to the preprocess2 code. sur$Result[sur$Question %in% c( "Better.availability", "Less.chemicals", "Recommended", "Not.recommended" ) & is.na(sur$Result)] <- 4 # If missing --> no change sur$Result[is.na(sur$Result)] <- 1 # Replace missing values with 1. That will produce 0 g/d. } return(Ovariable( output = sur, marginal = colnames(sur) %in% c("Fish", "Iter", "Question") )) } ) objects.store(jsp) cat("Ovariable jsp with actual survey data: every respondent is kept in data without sampling.\n")

Amount estimated from a bayesian model.

• Model run 24.5.2017 [24]

+ Show code- Hide code # This is code Op_en7749/modeljsp on page [[Goherr: Fish consumption study]] library(OpasnetUtils) jsp <- Ovariable( "jsp", dependencies = data.frame(Name = "fish.param", Ident = "Op_en7749/bayes"), formula = function(...) { require(MASS) require(reshape2) jsp <- lapply( 1:2, FUN = function(x) { mvrnorm(openv$N, fish.param$mu[,x], fish.param$Omega[[x]]) } ) jsp <- rbind( cbind( Fish = "Salmon", Iter = 1:nrow(jsp[[1]]), as.data.frame(jsp[[1]]) ), cbind( Fish = "Herring", Iter = 1:nrow(jsp[[2]]), as.data.frame(jsp[[2]]) ) ) jsp <- melt(jsp, id.vars = c("Iter", "Fish"), variable.name = "Question", value.name = "Result") jsp <- Ovariable(output=jsp, marginal = colnames(jsp) %in% c("Iter", "Fish", "Question")) return(jsp) } ) objects.store(jsp) cat("Ovariable jsp stored.\n")

Amount estimates directly from data rather than from a bayesian model.

• Initiation run 18.5.2017 [25]
• Initiation run 24.2.2018: sampling from survey rather than each respondent once [26]

+ Show code- Hide code # This is code Op_en7749/surveyjsp on page [[Goherr: Fish consumption study]] # The code produces amount esimates (jsp ovariable) directly from data rather than bayesian model. library(OpasnetUtils) jsp <- Ovariable( "jsp", dependencies = data.frame(Name="survey1", Ident="Op_en7749/preprocess2"), # [[Goherr: Food cunsumption study]] formula = function(...) { require(reshape2) sur <- survey1[ sample(1:nrow(survey1), openv$N, replace=TRUE, prob=survey1$Weighting) , c(157,1,3,158,162:178,81,82,131,132) # Removed, not needed:16,29,30? ] sur$Iter <- 1:openv$N #colnames(sur) #[1] "Row" "Country" "Gender" #[4] "Ages" "Baltic.salmon.n" "How.often.BS.n" #[7] "How.much.BS.n" "How.often.side.BS.n" "How.much.side.BS.n" #[10] "Better.availability.BS.n" "Less.chemicals.BS.n" "Eat.BH.n" #[13] "How.often.BH.n" "How.much.BH.n" "How.often.side.BH.n" #[16] "How.much.side.BH.n" "Better.availability.BH.n" "Less.chemicals.BH.n" #[19] "Eatfish" "Eatsalm" "Eatherr" #[22] "Recommended.BS" "Not.recommended.BS" "Recommended.BH" #[25] "Not.recommended.BH" "Iter" # Make sure that Row is kept separate from Iter because in the sampling version they are different. # sur contained columns Eat.fish, How.often.fish, Eat.salmon. Are these needed, as all other questions are melted? No # Columns 5-7 removed, so colnames list above does not match. colnames(sur)[5] <- "Eat.BS" colnames(sur) <- gsub("\\.n","",colnames(sur)) sur[22:25] <- sapply(sur[22:25], as.numeric) sur <- melt( sur, id.vars=c(1:4,26), variable.name="Question", value.name="Result" ) sur$Fish <- ifelse(grepl("BH",sur$Question),"Herring","Salmon") sur$Question <- gsub("\\.BS","",sur$Question) sur$Question <- gsub("\\.BH","",sur$Question) # The adjustments below probably should go to the preprocess2 code. sur$Result[sur$Question %in% c( "Better.availability", "Less.chemicals", "Recommended", "Not.recommended" ) & is.na(sur$Result)] <- 4 # If missing --> no change sur$Result[is.na(sur$Result)] <- 1 # Replace missing values with 1. That will produce 0 g/d. return(Ovariable( output = sur, marginal = colnames(sur) %in% c("Fish", "Iter", "Question") )) } ) objects.store(jsp) cat("Ovariable jsp with actual survey data: a respondent is sampled for each iteration.\n")

Initiate other ovariables

• Code stores ovariables assump, often, much, oftenside, muchside, amount.
• Model run 19.5.2017 [27]
• Initiation run 24.5.2017 without jsp [28]
• Model run 8.6.2017 [29]

+ Show code- Hide code # This is code Op_en7749/initiate on page [[Goherr: Fish consumption study]] library(OpasnetUtils) # Combine modelled survey answers with estimated amounts and frequencies by: # Rounding the modelled result and merging that with value in ovariable assump often <- Ovariable( "often", dependencies = data.frame( Name=c("jsp","assump"), Ident=c("Op_en7749/surveyjsp","Op_en7749/initiate") ), formula = function(...) { out <- jsp[jsp$Question == "How.often" , ] out$Value <- round(result(out)) out <- merge( assump@output[assump$Variable == "freq",], out@output ) out <- out[!colnames(out) %in% c("Value", "Variable", "Result","Question")] colnames(out)[colnames(out) == "assumpResult"] <- "Result" return(out) } ) much <- Ovariable( "much", dependencies = data.frame( Name=c("jsp","assump"), Ident=c("Op_en7749/surveyjsp","Op_en7749/initiate") ), formula = function(...) { out <- jsp[jsp$Question == "How.much" , ] out$Value <- round(result(out)) out <- merge( assump@output[assump$Variable == "amdish",], out@output ) out <- out[!colnames(out) %in% c("Value", "Variable", "Result","Question")] colnames(out)[colnames(out) == "assumpResult"] <- "Result" return(out) } ) oftenside <- Ovariable( "oftenside", dependencies = data.frame( Name=c("jsp","assump"), Ident=c("Op_en7749/surveyjsp","Op_en7749/initiate") ), formula = function(...) { out <- jsp[jsp$Question == "How.often.side" , ] out$Value <- round(result(out)) out <- merge( assump@output[assump$Variable == "freq",], out@output ) out <- out[!colnames(out) %in% c("Value", "Variable", "Result","Question")] colnames(out)[colnames(out) == "assumpResult"] <- "Result" return(out) } ) muchside <- Ovariable( "muchside", dependencies = data.frame( Name=c("jsp","assump"), Ident=c("Op_en7749/surveyjsp","Op_en7749/initiate") ), formula = function(...) { out <- jsp[jsp$Question == "How.much.side" , ] out$Value <- round(result(out)) out <- merge( assump@output[assump$Variable == "amside",], out@output ) out <- out[!colnames(out) %in% c("Value", "Variable", "Result","Question")] colnames(out)[colnames(out) == "assumpResult"] <- "Result" return(out) } ) assump <- Ovariable( "assump", ddata = "Op_en7749", subset = "Assumptions for calculations" ) effinfo <- Ovariable( # Effect of information "effinfo", dependencies = data.frame( Name=c("jsp","assump"), Ident=c("Op_en7749/surveyjsp","Op_en7749/initiate") ), formula = function(...) { away <- c("Value","Variable","Result","Question","assumpUnit","assumpSource","jspResult","jspSource") # Better availability of fish o1 <- jsp[jsp$Question == "Better.availability" , ] o1$Value <- round(result(o1)) o1 <- merge( assump@output[assump$Variable == "change",], o1@output ) colnames(o1)[colnames(o1)=="assumpResult"] <- "Better.availabilityResult" o1 <- Ovariable(name="Better.availability", output=o1[!colnames(o1) %in% away]) # Less chemicals o2 <- jsp[jsp$Question == "Less.chemicals" , ] o2$Value <- round(result(o2)) o2 <- merge( assump@output[assump$Variable == "change",], o2@output ) colnames(o2)[colnames(o2)=="assumpResult"] <- "Less.chemicalsResult" o2 <- Ovariable(name="Less.chemicals", output=o2[!colnames(o2) %in% away]) out <- (o1 + o2)@output return(Ovariable(output=out, marginal=colnames(out) %in% c("Iter","Fish"))) } ) effrecomm <- Ovariable( # Effect of recommendations "effrecomm", dependencies = data.frame( Name=c("jsp","assump"), Ident=c("Op_en7749/surveyjsp","Op_en7749/initiate") ), formula = function(...) { # Recommend more fish o1 <- jsp[jsp$Question == "Recommended" , ] o1$Value <- round(result(o1)) o1 <- merge( assump@output[assump$Variable == "change",], o1@output ) colnames(o1)[colnames(o1)=="assumpResult"] <- "Result" o1$Recommendation <- "Eat more" # Recommend less fish o2 <- jsp[jsp$Question == "Not.recommended" , ] o2$Value <- round(result(o2)) o2 <- merge( assump@output[assump$Variable == "change",], o2@output ) colnames(o2)[colnames(o2)=="assumpResult"] <- "Result" o2$Recommendation <- "Eat less" out <- rbind(o1, o2)[!colnames(o1) %in% c("Value","Variable","Question","assumpUnit","assumpSource","jspResult","jspSource") ] return(Ovariable(output=out, marginal=colnames(out) %in% c("Iter","Fish","Recommendation"))) } ) amountRaw <- Ovariable( "amountRaw", dependencies = data.frame(Name = c( "often", "much", "oftenside", "muchside", "assump" )), formula = function(...) { away <- c( "assumpUnit", "jspResult", "jspSource", "Variable", "Value", "Explanation" ) often <- often[ , !colnames(often@output) %in% away] much <- much[ , !colnames(much@output) %in% away] oftenside <- oftenside[ , !colnames(oftenside@output) %in% away] muchside <- muchside[ , !colnames(muchside@output) %in% away] assump <- assump[assump$Variable == "ingredient", !colnames(assump@output) %in% away] out <- (often * much + oftenside * muchside * assump)/365 # g /d return(out) } ) amount <- Ovariable( "amount", dependencies = data.frame(Name = c( "amountRaw", "effinfo", "effrecomm" )), formula = function(...) { out <- amountRaw * (1 + effinfo + effrecomm) result(out)[result(out)<0] <- 0 result(out)[result(out)>100] <- 100 return(out) } ) rm(wiki_username) objects.store(list=ls()) cat("Ovariables", ls(), "stored.\n")

Dependencies

The survey data will be used as input in the benefit-risk assessment of Baltic herring and salmon intake, which is part of the WP5 work in Goherr-project.

See also

• Useful information about Wishart distribution and related topics:
  • en:Scatter matrix
  • en: Wishart distribution
  • en:Chi-squared distribution

Related files

Goherr: Fish consumption study. Opasnet . [30]. Accessed 15 May 2025.