Final answer:
The statement claiming linear regression models are useless for non-linear patterns is false. With transformations, linear regression can fit certain types of non-linear patterns successfully, but other sophisticated models may be required if the relationship cannot be linearized. Correct option is b) False.
Step-by-step explanation:
The statement 'Linear regression models are of no use when fitting non-linear patterns' is False. Linear regression can be used to model a wide range of data, including some non-linear patterns, by transforming the data or the model. For example, transformations such as logarithmic, exponential, or polynomial can render a non-linear relationship into a linear one in terms of the transformed variables, thereby making linear regression applicable.
However, if the relationship between variables is non-linear and cannot be linearized by a simple transformation, more sophisticated models like polynomial regression or non-linear regression are then used. Linear regression is a procedure best suited for data that follows a linear trend, as indicated by a significant correlation coefficient (r) and a linear pattern in the scatter plot. When r is significant and the scatter plot shows a linear trend within the domain of observed x values, linear regression can be employed to predict the value of y. On the other hand, a small percentage of variation explained by the regression or a non-significant r indicates a weak linear relationship, implying that the regression line is not a good model for the data.