20241019 - hypothesis and prediction

references.bib

@@ -111863,6 +111863,17 @@ @Article{Enke2020
   url      = {https://doi.org/10.1093/qje/qjaa012},
 }
 
+
+@Article{Eastwell2014,
+  author  = {Eastwell, Peter},
+  journal = {Science Education Review},
+  title   = {Understanding hypotheses, predictions, laws, and theories},
+  year    = {2014},
+  number  = {1},
+  pages   = {16--21},
+  volume  = {13},
+}
+
 @Comment{jabref-meta: databaseType:bibtex;}
 
 @Comment{jabref-meta: saveActions:enabled;

research-methods.qmd

@@ -33,6 +33,18 @@ Although the convenience sample may afford a very large sample, the college stud
 Thus, the findings in our study may not *generalize* to NFL players—that is, what we learn in college students may not apply in the same way among NFL players.
 For instance, if we learn that consumption of sports drinks (compared to drinking only water) improves running speed among college students, that may not be the case among NFL players.
 
+## Research Questions, Hypotheses, and Predictions {#sec-hypothesisVsPrediction}
+
+A research question is a question that the investigator (you!) want to know the answer to.
+For example, a research question might be: "Does consumption of sports drink improve player performance?"
+A hypothesis is a proposed explanation.
+A prediction is "the expected result of a test that is derived, by deduction, from a hypothesis or theory" [@Eastwell2014, p. 17; archived at <https://perma.cc/8EX4-8JYN>].
+Here is an example of a hypothesis and the resulting prediction:
+
+> The present study evaluates whether consumption of sports drink improves player performance.
+> I <u>*hypothesize*</u> that consumption of sports drink leads football players to perform better in games because of greater endurance owing to restoration of electrolytes.
+> If the hypothesis is true, I <u>*predict*</u> that players who consume sports drink during a game will score more fantasy points than players who do not consume sports drink during the game.
+
 ## Research Designs {#sec-researchDesigns}
 
 There are three broad types of research designs:
@@ -58,7 +70,7 @@ An intervention that includes random assignment (e.g., to the intervention or co
 A randomized controlled trial (RCT) is an example of an experiment because it is an intervention with random assignment.
 
 For instance, we may be interested to evaluate whether players perform better (e.g., run faster) if they drink a sports drink compared when they drink only water.
-Our hypothesis might be that players will be expected to perform better when they drink a sports drink (compared to when they drink only water).
+Our hypothesis might be that players will be expected to perform better when they drink a sports drink (compared to when they drink only water), for the reasons specified in @sec-hypothesisVsPrediction.
 To this this research question and hypothesis, we might conduct an experiment by randomly assigning some players during practice to receive a sports drink and some players to receive only water.
 In this case, our independent variable is whether the player receives a sports drink.
 Our dependent variable might be their 40-yard dash time during practice.