William Ward
Assignment (no more than 2 pages single-spaced):
1. What does Abelson mean by “Statistics as a principled argument” in your own words?
By a principled argument, I believe that Abelson means that there should be a method to statistics. Not necessarily that it should be overly formalized or technical, but one with sound judgement. Abelson argues that good judgement with statistics in social science has been lost because there is considerable focus on null hypothesis testing (Abelson, 2). Statistical results that are shared should always be prefaced with a context and be provided with comparison groups. Also, it should never be the case for researchers to come up with an absolute conclusion once a model reaches the threshold of …show more content…
We should not just look at the P-value of an experiment and conclude that there is a relationship. We need to consider that our results are not absolute and there may be other reasons for why the relationships are observed. For example, can the phenomenon we are observing be better explained by an alternative explanation than what we are proposing? Could our results have been from chance with our data (i.e. measurement error or lucky sampling)? Is our data set strong enough to provide the results we are looking for? Can we prove this same finding through multiple …show more content…
Even by using very rigorous methodology there is still some level of error in our findings and will always be. It’s likely that if replicating the same study again with the same methods it may lead to the same conclusion but the strength of the conclusions may differ because of the error in sampling or measurement. Just using one study is not definitive and can easily be critiqued. When using statistical significance tests you will likely always find some difference when having a control and a logical manipulation (Abelson, 10). Because of these reasons it’s important to use the null hypothesis test as a guide, but not as the end all be
Assignment (no more than 2 pages single-spaced):
1. What does Abelson mean by “Statistics as a principled argument” in your own words?
By a principled argument, I believe that Abelson means that there should be a method to statistics. Not necessarily that it should be overly formalized or technical, but one with sound judgement. Abelson argues that good judgement with statistics in social science has been lost because there is considerable focus on null hypothesis testing (Abelson, 2). Statistical results that are shared should always be prefaced with a context and be provided with comparison groups. Also, it should never be the case for researchers to come up with an absolute conclusion once a model reaches the threshold of …show more content…
We should not just look at the P-value of an experiment and conclude that there is a relationship. We need to consider that our results are not absolute and there may be other reasons for why the relationships are observed. For example, can the phenomenon we are observing be better explained by an alternative explanation than what we are proposing? Could our results have been from chance with our data (i.e. measurement error or lucky sampling)? Is our data set strong enough to provide the results we are looking for? Can we prove this same finding through multiple …show more content…
Even by using very rigorous methodology there is still some level of error in our findings and will always be. It’s likely that if replicating the same study again with the same methods it may lead to the same conclusion but the strength of the conclusions may differ because of the error in sampling or measurement. Just using one study is not definitive and can easily be critiqued. When using statistical significance tests you will likely always find some difference when having a control and a logical manipulation (Abelson, 10). Because of these reasons it’s important to use the null hypothesis test as a guide, but not as the end all be