FINCH: THE MECCA PROJECT 



for the period reported. A computer model was 

 then constructed to scan the data base, using 

 for each run a given set of microconstituent 

 levels for the 52 kinds of fish reported in the 

 data base, and to calculate the average daily 

 intake of the microconstituent for each family 

 in turn, based upon its consumption by species. 

 From this a distribution of the estimated intake 

 levels of all the individuals in the survey was 

 calculated and printed out, with other data to 

 be described. This model was applied to calculate 

 mercury intakes from fish of individuals in 

 the sample on the basis of average levels of 

 mercury for the 52 kinds of fish in the data 

 base. Further computer runs were then made 

 using alternative input mercury levels. These 

 were calculated assuming the effects of different 

 regulatory guidelines upon the mercury levels 

 in the available supply. The program readouts 

 enabled the mercury intakes of the population 

 average, or of any given fraction of the popula- 

 tion, to be estimated under any of the input 

 conditions. By comparing results, the extent to 

 which different guidelines reduced the intakes 

 of mercury was estimated. This process is 

 described in detail in the following sections of 

 the report. 



Transformation of Survey 

 Information to Data Bases 



The estimates of consumption of kinds of 

 fish which comprised the data base were de- 

 veloped as follows: 



1. Records were selected from those families 

 who reported for a minimum of 12 periods 

 (6 mo). The number of reporting periods 

 was noted for each family. 



2. The purchases of each item by each family 

 during all reporting periods were totaled. 



3. These amounts were transformed into 

 corresponding figures for weights of each 

 kind of edible fish consumed by multiplying 



used in the survey and better understood by the average 

 consumer. For example, Ught meat tuna can legally 

 consist of several species, although in the United States 

 it commonly includes yellowfin tuna, Thunnus albacarcs; 

 bluehn tuna, Tlmiuuts ihynmis: and skipjack tuna, 

 Kutsiiwoniis pelaniis. Also for convenience of calculation, 

 some fish products, such as fish sticks, were identified 

 as kinds of fish in the data base rather than reassigning 

 them to component species. 



the total for each item by an appropriate 

 factor for that item. Table 1 lists these 

 factors and provides a note on their 

 derivation. 



4. The next step was to estimate the amounts 

 of fish items consumed in meals away 

 from home. The records showed for each 

 item the number of times it was consumed 

 in the reporting period, and how many 

 family members participated in each meal 

 out. From this, the total number of serv- 

 ings of each item for each family was 

 determined for the total reporting period. 



5. This number of servings was then con- 

 verted to the weight consumed, using a 

 factor for the item representing the average 

 weight in ounces usually provided in one 

 serving. Table 1 lists the factors used. 



6. The estimated weights for each kind of 

 sport fish consumed by each family was 

 obtained by multiplying the number of 

 events for each kind by the number in 

 the family and then by a factor represent- 

 ing the estimated average weight of the 

 edible portion of this kind usually con- 

 sumed by an average individual. These 

 factors shown in Table 1 are the same as 

 those used for meals away from home. 



7. The total weight of each kind of fish con- 

 sumed by each family was obtained by 

 adding the edible weights purchased to 

 the weights provided by meals away from 

 home and the sports catch consumption. 

 In a number of cases, several separately 

 reported items were aggregated to provide 

 total estimates of the consumption of a 

 given kind of fish from all sources. For 

 example, the total for the light meat tuna 

 category included canned light meat tuna, 

 tuna sandwiches, tuna salad, and tuna 

 pies. In this way over 150 different classes 

 of fish and products, many of them spe- 

 cialty items, were reduced to 52 kinds. 

 The finished data base displays the esti- 

 mated amounts of each of these 52 kinds 

 of fish consumed from all sources by each 

 family during the reporting period, the 

 length of reporting period, the number in 

 in the family, and coded socioeconomic 

 data provided by the original survey. 



617 



/ 



