Different sources of statistical and systematical uncertainties that can affect the projection factor of Cepheids. (1) Statistical uncertainty on the projection due to the uncertainty on the photometry (V, K, and A_V; blue bar). The resulting uncertainty on θ₀ is negligible (lower than the thickness of the bar). The red cross is the reference (see the text). (2) Same as panel 1, except that the value of θ₀ is shifted by 0.5 mas for clarity. The blue bar corresponds to the systematical uncertainty on p due to the choice of the SBCR in the literature, i.e., it corresponds to the standard deviation of p values in Fig. 10 or in Table 5. (3) Statistical uncertainties corresponding to the uncertainties on the coefficients of the S21 SBCR. (4) Range over which the projection factor varies (i.e., bias or systematics) depending on the choice of the spectral line in order to derive the radial velocity (among the list of 17 lines of Table 1). (5) Indicative systematical uncertainties on p and θ₀ due to a positive or negative 0.01 mag shift on the zero point of the SBCR. (6) Indicative systematical uncertainties on p and θ₀ due to a positive or negative 0.01 shift on the slope of the SBCR. (7) Offsets on p and θ₀ when E(B − V) changes from 0.1 to 0.2 mag (instead of 0.152). (8) Offsets on p and θ₀ when a shift in V from −0.1 to 0.1 mag is considered (due to a possible CSE) instead of 0 in the original fit (item 1). (9) Offsets on p and θ₀ when a shift in K of −0.1 mag is considered (due to a possible CSE) instead of 0.