Choose as large a sample as is practical. When you examine your evidence, you must determine when you have enough examples. Playing one guitar from a manufacturer may show you that the guitar has poor sound quality, but you would have to sample a few more guitars to conclude that the manufacturer produces poor quality guitars. Strengthen your inductive arguments by using as large a sample pool as you can -- within time constraints -- before drawing conclusions.
Identify possible exceptions to your argument. Inductive reasoning leads to likely conclusions, not irrefutable proof. You can strengthen your inductive arguments by acknowledging possible exceptions to your conclusions but focusing on the probability that your are right. Example: Most nurses are female. Bobbie is a nurse. You cannot simply conclude that Bobbie is female. You must acknowledge that you are making an educated guess. This strengthens your argument for listeners because you have taken away one of their probable objections to your line of reasoning.
Use random samples. If you only observe examples from a limited geographic area, demographic or other grouping, your conclusions may not be taking unusual circumstances into account. A person sitting in a restaurant watching injured pedestrians hobbling by on crutches might conclude traffic is dangerous in that city before he realizes he is next to an ambulatory care facility. Make your samples as diverse as possible, drawn randomly from the sample pool.
Question your analogies. An analogy assumes something is true because two samples are similar. Example: Bill is tall and wise. James is tall, so he must be wise. This conclusion is not warranted, and is based on an analogy; because Bill and James share one similarity; they share others. Analogies can destroy your credibility if they jump to conclusions based on mere similarity. Avoid analogies as much as possible, and use them only as starting places for an inquiry, not as means to a conclusion.