Page 165 - Data Science class 10
P. 165

There are several types of sample selection bias, including pre-screening bias, self-selection bias, exclusion bias,
            and observer bias.
            Example of Sample Selection Bias
            For example, medical studies, that recruit participants directly from clinics, are bound to miss all those who don’t
            attend these clinics or seek care during the study.
            Researchers and study organizers have the responsibility to ensure that the results of their studies are accurate,
            relevant, and do not include any prejudice that might result in incorrect conclusions. One way to do this is to
            structure the study based on a method that supports a random sample selection process.
            While in principle this may sound simple enough, in reality the researcher will need to exercise caution in order
            to avoid sample selection bias. Additionally,  the  study organizer  may be faced  with  restrictions beyond  their
            control that make it challenging to realise a random sample. For example, there may be a lack of participants or
            inadequate funding for the project.
            Ensuring that the sample being examined is randomly chosen, the researcher should identify the various subgroups
            within the population.  They should then analyse the sample  to determine if these subgroups are adequately
            represented in the study.

            In some cases, the researcher may find that certain subgroups are either over-represented or under-represented
            in their study. The researcher can now use bias correcting techniques. One method is to assign weights to the
            misrepresented subgroups in order to statistically correct the bias. This weighted average considers the relative
            importance of each subgroup and can lead to results that more accurately reflect the study population's actual
            demographics.

            3.3.2. Linearity Bias

            Linearity Bias is the belief that changing one quantity would automatically result in a corresponding change in
            another. Unlike Selection Bias, Linearity Bias is a cognitive bias; it's produced not through some statistical process,
            but rather via how we mistakenly interpret the world around us.

            For example, let us take the case of relationship between fuel efficiency and engine displacement (i.e. engine size)
            of automobiles?

            Automobile engineering tells us that, as engines become larger, their fuel efficiency decreases. The majority of
            people believe that the relationship between engine displacement and fuel economy is a straight line. It is, at first,
            but real data tells a more complex story for the full picture, as shown in following figure:

                                                    Fuel efficiency vs. Engine size



                                    40                                               class
                                   Highway fuel efficiency (mpg)  30                    2seater


                                                                                        compact
                                                                                        midsize
                                                                                        minivan
                                                                                        pickup
                                                                                        subcompact
                                                                                        suv
                                    20





                                            2      3       4      5       6      7
                                                       Engine displaement (L)





                                                                                          Identifying Patterns   163
   160   161   162   163   164   165   166   167   168   169   170