Table 4. Linear mixed effect model of five geographical and socioeconomic variables in which attended school and the place of residence were selected as random effects (a sample of 1,708 males for which all data were available).
Individual counties are ranked according to the latitude of their capitals.
| Parameter | Estimate (height values in cm) | 95% Confidence interval | p-value |
|---|---|---|---|
| Intercept (constant) | 180.6 | 179.7, 181.6 | <0.001 |
| School Type: Gimnazija | 0.56 | −0.24, 1.37 | 0.17 |
| School Type: Mixed | 0.76 | −0.19, 1.71 | 0.12 |
| School Type: Other | 0a | ||
| Age: 17.0–17.9 years | 0.30 | −0.68, 1.27 | 0.55 |
| Age: 18.0 + years | 0a | ||
| County: DUBROVNIK-NERETVA | 2.81 | 1.55, 4.07 | <0.001 |
| County: SPLIT-DALMACIJA | 3.19 | 1.99, 4.38 | <0.001 |
| County: ŠIBENIK-KNIN | 2.89 | 1.55, 4.24 | <0.001 |
| County: ZADAR | 2.32 | 1.07, 3.58 | <0.001 |
| County: ISTRA | 0.32 | −1.12, 1.77 | 0.66 |
| County: LIKA-SENJ | 0.45 | −1.29, 2.19 | 0.61 |
| County: PRIMORJE-GORSKI-KOTAR | 1.29 | 0.04, 2.55 | 0.044 |
| County: KARLOVAC | 0a | ||
| Urban residence: NO | −0.66 | −1.34, 0.02 | 0.057 |
| Urban residence: YES | 0a | ||
| University education of parents: BOTH | 0.20 | −0.78, 1.19 | 0.69 |
| University education of parents: ONE | −0.12 | −0.87, 0.64 | 0.76 |
| University education of parents: NEITHER | 0a |
Notes.
Significant values are highlighted in red. The sum of values from individual categories can predict mean male height for the particular combination of factors, relative to the intercept (180.7 cm). For example, males from the county of Split-Dalmacija (+3.19 cm), aged 18 + years (0 cm), from elite high schools (gimnazija) (+0.56 cm), urban localities (0 cm), and with university-educated parents (+0.20 cm) would be 184.6 cm tall.
a
This parameter represents a standard within the particular category and is set to zero.