Observations as a research method requires the researcher to watch subjects in a natural or artificial setting and record their behaviour. One advantage of naturalistic observations is that they take place in the participant's everyday environment, gaining highly valid information of their natural behaviour. However, if the observer is overtly watching then it can lead to the observer effect as participants may change their behaviour, even without realising. Additionally, structured observations may skew behaviour as the participants are in an artificial environment that may not be representative of their real life. Another advantage is that observations are highly reliable if tallying and time sampling techniques are used because more than one observer can check for inter-observer reliability. A weakness of this method is that important behaviour may be missed if it does not fit in with the categories invited prior to the study.