## Table 3 |
||

Parameter estimates of the alternative bmi structural model specifications (t statistics
in parentheses) |
||

Independent variables |
Instrumental variables regression coefficients (t-statistics) |
Least squares regression coefficients (t-statistics) |

INTERCEPT | 27.96 | 28.21 |

(82.41)** | (83.45)** | |

AGE^{a} |
0.22 | 0.22 |

(16.57)** | (16.66)** | |

AGE_SQ^{a} |
−0.00 | −0.01 |

(−2.71)** | (−2.89)** | |

EDUC | −0.31 | −0.33 |

(−13.04)** | (−13.88)** | |

MARRIED | 0.33 | 0.38 |

(2.62)** | (2.97)** | |

BMIYR | 0.09 | 0.10 |

(7.54)** | (8.11)** | |

SIBBMI | 2.34 | 2.35 |

(35.82)** | (35.75)** | |

RESWALK | −0.24 | −0.00 |

(−3.23)** | (−0.08) | |

Adjusted R^{2} |
0.16 | 0.16 |

F Statistic | 268.52** | 267.90** |

**p < .05.

^{a}Age is centered before creating age squared so as to avoid any possible collinearity
problems associated with entering both age and age squared as regressors.

Zick * et al.*

Zick * et al.* *International Journal of Behavioral Nutrition and Physical Activity* 2013 **10**:27 doi:10.1186/1479-5868-10-27