Teachers with content-specific training in the subject of math has been found to be more effective in increasing math achievement than are teachers without such credentials. Druva and Anderson (1983), in a meta-analysis of studies looking at the relationship between teacher background and students' science achievement, found that teachers' number of science courses taken, experience teaching biology (for biology students), and attendance at academic institutes were positively correlated with students' achievement outcomes in higher level science courses. Goldhaber and Brewer (1997), in an analysis of NELS 8th and 10th grade data, found that higher math achievement is found in students with teachers with math training, after accounting for background factors. The sample was made up of 35,149 public school students from 638 schools with 2,245 different mathematics teachers.