The head of the accounting department randomly examined some accounting entries and was upset with the high proportion of incorrect invoices. He instituted a new system to keep the proportion of bad invoices below 0.1 %.

A year later, 10,000 invoices were randomly examined and six were found to be incorrect. Can this manager reject the null hypothesis that the proportion of bad invoices is 0.1 %? Do a 5 % level of significance test.