Factor loading ranges from -1 to 1. Loadings that are closer to -1 or 1 signifies higher association between the observed variable and the factor, whereas loadings closer to 0 show poor association. In the table below, Hair et al. (2009) recommend cutoff for factor loadings relative to sample size. For example, for a sample size of 250, a factor loading of 0.35 is needed for significance. Note that negative factor loading means the item/question is inversely worded relative to other items measuring the same factor with higher loadings. These items need to be reverse coded to be coded. Alternatively, you can take the absolute of the value.