I agree with what the others have written. In qualitative research, there are more threats to validity. The researcher is the instrument, and researchers are human. Since we are not crunching numbers, there is more opportunity to be subjective. When interviewing or observing, we can get tired or distracted. We also have our own assumptions to get past. These are often the biggest threat. We need to keep an open mind, and that is hard to do!
Our own personal paradigms are perhaps the biggest obstacle to research. Research should be objective, and if we are not very careful of our own biases, we might overlook research that is relevant but contrary to what we are trying to prove or achieve. Along those lines, we must be mindful that others' research which we incorporate into our own might be biased as well.
This depends a lot on what sort of research you are doing. The first post assumes you're doing scientific research. If you are doing research in social science, you need to check for confounding variables that might be messing up the relationship between your independent variable and your dependent variable.
For example, let's say you want to research the connection between tax rates and how strong the economy is. Let's say you find that tax rates went down and the economy went into a recession. Can you say the taxes caused that? Maybe, but you need to make sure nothing else was going on that caused the recession. You have to be sure that there are no factors other than your independent variable that could be causing the changes in your dependent variable.
First of all, if any equipment is being used, it must be checked for accuracy. For example, if weighing something in the experiment, did the experimenter make sure the digital scale was on 0 before weighing the item? Also, many trials must be performed and a large sample must be used, to get more valid results. Is the experiment based on valid mathematical principles? This is important to be able to weed out any inconsistencies that may arise. The data must be verified by the experimenter. After many trials erroneous data can be identified and the experimenter may want to re-do or redesign her experiment. Can the results be duplicated by other scientists? If many researchers arrive at the same result, the experiment would seem to be valid. However, if the results can't be replicated, perhaps a new hypothesis needs to be tested.