Slide 10 this slide lays out the scientific method

This preview shows page 4 - 6 out of 53 pages.

SLIDE 10 This slide lays out the scientific method. Sociology, as opposed to psychology, is typically non-causal, regardless of whether data is qualitative (meaning based) or quantitative (numbers based). Sociologists don’t often perform experiments, simply because of the sheer number of variables that cannot be accounted for when conducting a study on thousands of people. Thousands of people cannot be brought into a lab; it’s impractical and expensive. Instead, sociologists might use surveys to gather large amounts of data. Without getting overly technical, the way that most quantitative data is analyzed is using descriptive statistics (mean, median, and mode), measures of dispersion (looking at data ranges), and various modeling. The most commonly used model is called an OLS
(ordinary least squares) regression, and essentially plots a line of best fit across all cases. Then, just like when someone has a musical recording on a soundboard and can play with the treble or the bass by moving a lever up and down and listening to the result, sociologists analyze data in a similar way—turning one variable ‘all the way up’ to see what happens, or turning other variables off in the model. This lets us zero in on what factors are really causing the outcome versus which ones are just ‘background noise.’ SLIDES 11-14 Correlation is not causation. This is an important idea, but also one that is sometimes counter-intuitive. Correlation means that two things may share some directional relationship. That is, if one thing increases, another increases; if one decreases, another decreases. Thus, two variables are literally Co-Related. Causation on the other hand means one thing has directly caused another to happen, like cause and effect. Causally related (not *casually related, which is a common spelling error) items mean one has had a direct influence on the outcome of another. The reason why it’s important to understand the difference here is because lots of things are correlated, but not necessarily causally. Thus, you’ll see in some of these slides that the number of films that Nicolas Cage has appeared in IS actually correlated with the number of drowning deaths in swimming pools over time (this is true!). And while it’s tempting to believe that some people simply could not watch National Treasure even one more time and decided to end it all, we know that these two phenomena are statistically not causally related. Something very similar happens for example when it comes to children’s vaccines like the MMR (Measles, Mumps, Rubella Vaccine) and the onset of autistic symptomology. Autism is generally screened for by pediatricians at 18 months old. Most children are screened between a year and a half and five years old. Most children also receive their first MMR vaccine between twelve and fifteen months old. Autism screenings (and potential diagnoses) are statically correlated with the administration of the MMR vaccine due to the age ranges involved. The causal connection between the two phenomena however has been thoroughly debunked. The scientific consensus, based on

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture