The three major types of research designs in Psychology
research are descriptive, correlational, and experimental (“Introduction
to Psychology”, 2015). Descriptive research provides information of what
is occurring at a given time. It allows the development of questions for
further study. However, it does not assess relationships among variables and
therefore is not suitable for this research study.
Correlational research is used to discover relationship among two or more variables
but it cannot draw inferences about the causal relationships between the
research method is used when researcher manipulate an independent variable and
evaluate the impacts on the other variables (“Introduction
to Psychology”, 2015). This method can
identify the causal relationships among the variables but researcher can only
manipulate one variable at one time.
This research scenario used the experimental research
method. The researcher manipulates the independent variable, “name”, and
evaluation the response of the scientists of the Australian universities, the
dependent variable. In this research study, descriptive research method is not
appropriate because it does not provide information on the causal
relationship. Correlational research is
not relevant because it does not show the causal relationship between the
can be spotted using histograms and boxplots (Field, 2013). Histogram will show
whether the data is normally distributed evenly. If the curve is squashed towards one side and
if there is an isolated blocked at the tail of the histogram, it is possible
that there are data points falling away as the extremes.
is to inspect the boxplot of SPSS. Outliers are displayed as little circles
with a ID number attached in the boxplot.
b) Winsorizing data is one of the methods to
eliminate outliers (Field, 2013). It involves bringing the outliers down to a
specified value, so they are closer to within the normal distribution curve. One
disadvantage of winsorizing is that it makes you feel like cheating because you
change data to improve accuracy of your result and may in turn bias your result
(Field, 2013). Further, changing data also feels like a self-fulfilling act
because the data changed originate from the data set itself and obviously the
overall analysis will support it. However, if the score changed is originally
very unrepresentative of the sample and biases your result, changing it will
improve the accuracy of the study.
Significant result from ANOVA indicating
that suitability for STEM employment is influenced by candidates’ gender. For
the reader to better appreciate the importance of the findings, effect size of
the result should be measured (APA Manual
(Publication manual of the American Psychological Association), 2010).
Follow-up tests (post hoc
analysis) are usually required if the ANOVA result is statistically significant.
This is performed to contrast and compare the different groups. The most common
post hoc test is the Tukey’s honestly significant difference (HSD). It compares
all possible pairs of means within the data (Allen, Bennett & Heritage,
2014). Based on the compared result, we can tell whether the masculine name or
the feminine name or the unisex name has the strongest effect. For pairwise
comparisons, Tukey’s HSD is considered the most appropriate especially when confidence
intervals are needed or sample sizes are not equal (Multiple-Comparison Procedures, 2016).