Inference is the logical process of moving from an indicator or observation to a conclusion or general rule. Inference means reading all the clues and making your guess. Inference is studied within several different fields. Inference is the act or process of deriving a conclusion based solely on what one already knows. Statisticians have developed formal rules for inference from quantitative data. Inference may or may not be a correct one. Correctness is not what makes the difference between observation and inference. Human inference is traditionally studied within the field of cognitive psychology.
In logic studies, one progresses from certain premises to certain conclusions. AI researchers develop automated inference systems. AI systems for automated logical inference are popular research topics and have industrial applications under the form of expert systems.
It is important to distinguish between observations and inferences. The colors and color changes, the temperature and temperature changes, the smells that you may come across in this lesson and throughout this course are directly observed and they can be classified as observations. When you do something with that observation, like draw a conclusion or offer an explanation or decide that a chemical reaction occurred, then you are making an inference.
Statistical tools or techniques used to draw inference about a population on the basis of research evidence from a sample. For example, estimating the frequency of value for a particular variable within a population. This is found commonly in reports of public opinion polls when it is noted that a sample of this size is accurate to within +/- 3.2%, 19 times out of 20.