Skip to main content

Table 2 Finalised linear mixed models (LMM) of food spatial memory performance across cultures

From: Human spatial memory is biased towards high-calorie foods: a cross-cultural online experiment

USA

Ba

95% CI

p

ηp2 b

90%CI ηp2

Model 1: Best-fitting LMMc

 Intercept

442.98

292.27 – 593.69

 < .001*

-

-

 Caloric Density (High – Low)

-99.23

-197.19—-1.28

.047*

0.06

0.0004 – 0.16

 Taste (Sweet – Savoury)

-5.32

-22.58 – 11.95

.546

-

-

 Ethnicity

-

-

.041*

0.16

0.02 – 0.63

  Caucasian—Black

-135.70

-222.10—-49.30

.003*

-

-

  Black—Asian

191.45

66.31 – 316.59

.003*

-

-

  Black—Latino

168.86

13.50 – 324.21

.034*

-

-

 Caloric Density*Education

-

-

.083

-

-

  Low*Education

-21.18

-43.16 – 0.80

.059

-

-

Model 2: Best-fitting LMM with exploratory covariates (i.e. BMI, Healthy Eating Goals, and Encoding Time)c

 Intercept

244.44

-10.97 – 499.86

.103

-

-

 Caloric Density (High – Low)

-102.22

-201.09—-3.33

.043*

0.06

0.001 – 0.17

 Taste (Sweet – Savoury)

-4.52

-22.01 – 12.96

.612

-

-

 Ethnicity

-

-

.047*

0.17

0.007 – 0.65

  Caucasian—Black

-116.19

-197.69—-34.69

.006*

-

-

  Black—Asian

166.22

49.04 – 283.41

.006*

-

-

  Black—Latino

196.05

49.71 – 342.39

.010*

-

-

 Caloric Density*Education

-

-

.128

-

-

 BMI

5.12

1.02 – 9.21

.015*

0.09

0.01 – 0.23

 Healthy Eating Goals

2.80

-17.08 – 22.68

.779

-

-

 Encoding Time

-0.003

-0.01—-0.001

.001*

0.02

0.01 – 0.05

Japan

Model 1: Best-fitting LMM (with and without exploratory covariates BMI, Healthy Eating Goals, and Encoding Time)c

 Intercept

387.73

314.22 – 461.23

 < .001*

-

-

 Caloric Density (High – Low)

-69.85

-124.53—-15.17

.161

-

-

 Taste (Sweet – Savoury)

-7.89

-33.22 – 17.43

.251

-

-

 Desirability

-0.59

-1.01—-0.17

.006*

0.05

0.001 – 0.01

 Sex (Males – Females)

70.01

3.28 – 136.75

.040*

0.06

0.002 – 0.17

 Caloric Density*Taste

-

-

.044*

0.003

0.0001 – 0.01

  High-Savoury – Low-Savoury

-40.41

-76.14—-4.68

.025*

-

-

 Caloric Density*Occupation

-

-

.045*

0.09

0.002 – 0.32

 Low-Employed – Low-Unemployed

-90.93

-168.89—-12.96

.023*

-

-

Combined (USA + Japan + Netherlands)

Model 1: Best-fitting LMM d

 Intercept

1.83

1.65 – 2.00

 < .001*

-

-

 Country

-

-

 < .001*

0.08

0.09 – 0.22

  USA – Netherlands

0.29

0.19 – 0.39

 < .001*

-

-

  Japan—Netherlands

0.25

0.16 – 0.35

 < .001*

-

-

 Caloric Density (High – Low)

-0.03

-0.06—-0.002

.036*

0.01

0.0003 – 0.02

 Taste (Sweet – Savoury)

0.03

0.01 – 0.05

.009*

0.001

0.0001 – 0.001

 Sex (Males – Females)

0.09

0.03 – 0.16

.007*

0.01

0.002 – 0.03

 Age

0.007

0.005 – 0.01

 < .001*

0.08

0.05 – 0.12

 Education

-0.04

-0.07—-0.21

 < .001*

0.03

0.01 – 0.05

 Desirability

-0.0005

-0.001—-0.0001

.015*

0.001

0.0001 – 0.001

Model 2: Best-fitting LMM with exploratory covariates (i.e. BMI and Task/Encoding Time) d

 Intercept

1.68

1.42 – 1.94

 < .001*

-

-

 Country

-

-

 < .001*

0.09

0.1 – 0.23

  USA – Netherlands

0.29

0.19 – 0.39

 < .001*

-

-

  Japan—Netherlands

0.29

0.19 – 0.39

 < .001*

-

-

 Caloric Density (High – Low)

-0.03

-0.06—-0.002

.037*

0.01

0.0002 – 0.02

 Taste (Sweet – Savoury)

0.03

0.007 – 0.05

.009*

0.001

0.0001 – 0.002

 Sex (Males – Females)

0.09

0.02 – 0.15

.012*

0.01

0.001 – 0.03

 Age

0.007

0.005 – 0.01

 < .001*

0.08

0.04 – 0.1

 Education

-0.04

-0.06—-0.02

.001*

0.02

0.01 – 0.05

 Desirability

-0.0005

-0.001—-0.0001

.015*

0.001

0.0001 – 0.001

 BMI

0.005

-0.002 – 0.01

.148

-

-

 Task (Encoding) Time

-3.7 × 10–6

-5.9—-1.5 × 10–6

.001*

0.001

0.0002 – 0.002

  1. *Significant at α = 0.05
  2. aThe change in food spatial memory accuracy (D; pixels) associated with a one unit change in the predictor, with other model predictors held constant. Categorical predictors with more than 2 groups (e.g. Ethnicity), as well as interaction effects, are represented by multiple B coefficients – one for each significant group difference as revealed in Fisher’s LSD post-hoc tests
  3. bThe proportion of variance in food spatial memory that is explained by the predictor, after accounting for effects of other model predictors
  4. cA linear mixed model with food spatial memory accuracy (D; pixels) as the dependent variable
  5. dA linear mixed model with log10 (y + 1) transformed food spatial memory accuracy (D; pixels) as the dependent variable