Correlational finding on Happiness and subject: What one eats

StudyFord et al. (2013): study ZZ Anglo-America 2002
TitleIntake of Meditteranean Foods Associated with Positive and Low Negative Affect
SourceJournal of psychosomatic research, 74 (2013) 142 - 148
DOIdoi: 10.1016/j.jpsychores.2012.11.002
Public30+ aged, Adventist church atendees, North America 2002-2007
SampleProbability sample (unspecified)
Respondents N =9255

Author's labelFish
Page in Source 143-146
Our classificationWhat one eats
Selfreported  intake of fish during the last year
1  never or rarely
2  1-3 times per month
3  1-6  times per week
4  1-4 times per day
Observed distributionMmales: 7,6, females 7,4
Assessed usung Food Frequency Questionnaire (FFQ), 
which contains a list of 200 food items.

Fish types are: white fish (cod, salt fish, sole, 
haddock catfish), or halibut, snapper, tuna salad, 
salmon and una casserole.

Intake expressed in standard servings in grams
Food intake was energy adjusted using the residual 

Observed Relation with Happiness
A-BW-cy-mq-v-5-dBeta=+ p < .03
positive affect:  Beta =  +.03 (026)
negative affect:  Beta =  +.02 (301)

Beta's controled for:
- age
- gender
- ethnicity
- frequency of vigorous physical activity
- hours sleep
- total energy intake
- use of alcohol

Significant among males only

Appendix 1: Happiness measures used
CodeFull Text
A-BW-cy-mq-v-5-dSelfreport on 10 questions.

This scale consists of a number of words that describe different feelings and emotions. As I read each item, I ask that you refer to your response card an d tell me to wgat extent you have felt in that way during the past year :
A excited
B enthusiastic
C alert
D inspired
E determined
F distressed
G scared
H upset
I nervous
J ashamed

Answer options:
1 not at all
2 a little
3 somewhat
4 quite a bit
5 very much

Negative affect score (NAS): F to J
Positive affect score (PAS): A to G
Affect Balance Score (ABS): PAS - NAS

Name: Shortened Watson's PANAS; Kiercher version

Appendix 2: Statistics used
Type: test statistic.

Measurement level: Correlates: all metric, Happiness: metric.
Range: [-1 ; +1]

beta > 0 a higher correlate level corresponds to a higher happiness rating on average.
beta < 0 a higher correlate level corresponds to a higher happiness rating on average.
beta = 0 no correlation.
beta = + 1 or -1 perfect correlation.
Ruut Veenhoven, World Database of Happiness, Collection of Correlational Findings, Erasmus University Rotterdam.