Stat 1023 Assignment 2 Example Assignment 2 - comments .docx

Stat 1023 Assignment 2
Example Assignment 2 - comments
This example (on the next page) is based on the media story,
“The 6 most common regrets men and
women have after sex” (http://globalnews.ca/news/991303/the-
6-most-common-regrets-men-and-
women-have-after-sex/). The original research report that
describes the study is called, “Sexual regret:
Evidence for evolved sex differences”; I’ve posted the report as
a separate file alongside this example in
case you wish to look at it. If you do, you’ll notice that the
report describes three (3) separate studies.
The original media story only discussed the results and
conclusions of Study 3. Consequently, my version
of the media story also only deals with Study 3. Finally, for
creative purposes, I’ve created some
imaginary quotes from one of the authors; I certainly didn’t
interview the author—these simply
represent what I think the author might say based on the
information in the original research report.
I’m using the ‘abbreviated title’ of ‘Regrets’—notice that this is
how I named my file as well (otherwise, I
would have lost 3 marks!).
Remember that the purpose of this example is to show you the
type of detail and structure that your
assignment should demonstrate. The application/description of
course material in this example may
not be correct; it is simply an example of format and level of

detail. While you read this example, refer
back to the ‘Steps to complete this assignment’ so you
understand how those steps translate into the
completed assignment.
http://globalnews.ca/news/991303/the-6-most-common-regrets-
men-and-women-have-after-sex/
http://globalnews.ca/news/991303/the-6-most-common-regrets-
men-and-women-have-after-sex/
Feeling regretful about your last sexual encounter? Your gender
might matter!
At a singles mixer last weekend where you missed the
opportunity to get to know someone of the opposite sex a
little more ‘intimately’? Are you feeling a little remorseful
about your missed opportunity, and wondering
whether the other person feels the same? Turns out, they
probably don’t! According to a recent research study,
males and females have different regrets when it comes to
casual sex encounters.
Researchers from UCLA and the University of Texas explored
differences in how men and women respond to
participating in and passing up opportunities for casual sex.
“We expected males to regret passing up an
opportunity for casual sex more than females, and females to

regret engaging in casual sex more than
males….our data suggest we were right,” explains Martie
Haselton, co-author of the study.
Haselton further explains that these differences might result
from evolutionary differences between the genders.
That is, with females taking on the bulk of reproductive effort
(think about those nine months a woman must
spend pregnant if a sexual encounter results in conception!),
casual sex opportunities that are acted on could
come with a energy- and time-expensive cost. Males, on the
other hand, have an evolutionary history that
encourages more sexual encounters; consequently, males should
worry about lost potential for reproduction in
the form of ‘missed’ sexual encounters.
Does this research suggest that—as a female or male—you’ll
always regret casual sex encounters according to
your gender? Not necessarily: Haselton and colleagues based
their conclusions on an internet survey that relied
on visitors to a news site to click on a survey asking them to
think about their level of regret (from “I’m glad I
did it” to “I regret it very much”) after their last time they had
casual sex with someone, or, passed up a chance

to have casual sex with someone. Over 24000 individuals from a
diverse background (in terms of age, sexual
orientation, educational background, and current relationship
status) did respond to the survey—a very large
sample size. Even though they achieved diversity in their
respondents, Haselton points out that they had little
control over who was responding and whether they were being
completely honest (we’re talking about sexual
regret here!).
“We recognize that internet surveys come with their own
limitations—who knows who’s answering the survey!
But, we still feel confident in the data collected,” elaborates
Haselton. He points to the fact that survey was
anonymous, so people responding should have no pressure to be
dishonest about their sexual regrets and
experiences.
Finally, Haselton emphasizes that everyone will feel differently
about their casual sex opportunities: “Some of
us regret our actions deeply (and hopefully learn from them),
some of us don’t. Those differences are okay and

what make us unique! As evolutionary psychologists, our
research team is simply interested in how our
evolutionary history might influence these differences.”
Original Research Report:
Galperin, A., Haselton, M.G., Frederick, D.A., Poore, J., von
Hippel, W., Buss, D.M., and G.C. Gonzaga. 2012.
Sexual regret: Evidence for evolved sex differences. Archives
of Sexual Behavior 42(7): 1145-1161. Doi:
10.1007/s10508-012-0019-3
Knee osteoarthritis has doubled in prevalence since the
mid-20th century
Ian J. Wallacea, Steven Worthingtonb, David T. Felsonc, Robert
D. Jurmaind, Kimberly T. Wrene, Heli Maijanenf,
Robert J. Woodsg, and Daniel E. Liebermana,1
aDepartment of Human Evolutionary Biology, Harvard
University, Cambridge, MA 02138; bInstitute for Quantitative
Social Science, Harvard University,
Cambridge, MA 02138; cClinical Epidemiology Unit, Boston
University School of Medicine, Boston, MA 02118;
dDepartment of Anthropology, San Jose State
University, San Jose, CA 95192; eDepartment of Anthropology,
University of Tennessee, Knoxville, TN 37996; fLaboratory of
Archeology, University of Oulu,
Oulu 90014, Finland; and gBattelle Memorial Institute, Natick,
MA 01760

Edited by Osbjorn Pearson, University of New Mexico,
Albuquerque, NM, and accepted by Editorial Board Member C.
O. Lovejoy July 12, 2017 (received for
review March 7, 2017)
Knee osteoarthritis (OA) is believed to be highly prevalent
today
because of recent increases in life expectancy and body mass
index
(BMI), but this assumption has not been tested using long-term
historical or evolutionary data. We analyzed long-term trends in
knee OA prevalence in the United States using cadaver-derived
skeletons of people aged ≥50 y whose BMI at death was docu-
mented and who lived during the early industrial era (1800s to
early 1900s; n = 1,581) and the modern postindustrial era (late
1900s to early 2000s; n = 819). Knee OA among individuals
esti-
mated to be ≥50 y old was also assessed in archeologically
derived
skeletons of prehistoric hunter-gatherers and early farmers
(6000–
300 B.P.; n = 176). OA was diagnosed based on the presence of
eburnation (polish from bone-on-bone contact). Overall, knee
OA
prevalence was found to be 16% among the postindustrial
sample
but only 6% and 8% among the early industrial and prehistoric
samples, respectively. After controlling for age, BMI, and other
variables, knee OA prevalence was 2.1-fold higher (95% confi-
dence interval, 1.5–3.1) in the postindustrial sample than in the
early industrial sample. Our results indicate that increases in
lon-
gevity and BMI are insufficient to explain the approximate dou-
bling of knee OA prevalence that has occurred in the United
States

since the mid-20th century. Knee OA is thus more preventable
than is commonly assumed, but prevention will require research
on additional independent risk factors that either arose or have
become amplified in the postindustrial era.
arthritis | aging | obesity | mismatch disease | evolutionary
medicine
Osteoarthritis (OA) is the most prevalent joint disease and
aleading source of chronic pain and disability in the United
States (1) and other developed nations (2). Knee OA accounts
for more than 80% of the disease’s total burden (2) and affects
at
least 19% of American adults aged 45 y and older (3). Sub-
stantial evidence indicates that knee OA is proximately caused
by the breakdown of joint tissues from mechanical loading (4)
and inflammation (5), but the deeper underlying causes of knee
OA’s high prevalence remain unclear and poorly tested, hin-
dering efforts to prevent and treat the disease. Two recent
public
health trends, however, are commonly assumed to be dominant
factors (6, 7). First, because knee OA’s prevalence increases
with
age (8), the rise in life expectancy in the United States since the
early 20th century is thought to have led to high knee OA levels
among the elderly, with the presumption that, as people age,
their senescing joint tissues accumulate more wear and tear
from
loading (9). Second, high body mass index (BMI) has become
epidemic in the United States in recent decades and is a well-
known risk factor for knee OA (8), probably because of the
combined effects of joint overloading and adiposity-induced in-
flammation (10). Whether increases in longevity and BMI are
responsible for current knee OA levels has never been tested,
but
this assumption has led many to view the disease’s high preva-

lence as effectively unpreventable, since aging is untreatable,
and
the high BMI epidemic is intractable (8, 11).
One underused yet potentially powerful way to identify and
assess the risk factors responsible for current knee OA levels is
to
examine long-term changes in the disease’s prevalence by com-
paring contemporary with historic and prehistoric populations
(12). Epidemiological studies of present day populations are
valuable but are limited in their ability to analyze risk factors
that
are now pervasive but used to be less common. It is difficult to
find large samples of living Americans whose lifestyles,
including
physical activity levels and diet, resemble those of past genera-
tions. Although many variables cannot be measured and thus
controlled in epidemiological studies of people living in the
past,
a major benefit of analyzing populations over historical and
evolutionary time is to assess known risk factors under different
environmental conditions and thus bring to light the effects of
risk factors that might not be apparent or testable in modern
populations alone. Furthermore, although knee OA is known to
be ancient (12), we know very little about changes in its preva-
lence over time. Low levels of knee OA have been reported for
some historic and prehistoric populations (13–17), suggesting
that the disease’s prevalence has recently increased, but these
studies used different diagnostic criteria than those used to di-
agnose knee OA in living patients, used samples composed
mostly
of younger individuals, and did not account for BMI,
complicating
comparisons with modern epidemiological data.
Here, we investigate long-term trends in knee OA prevalence

in the United States and evaluate the effects of longevity and
Significance
Knee osteoarthritis is a highly prevalent, disabling joint disease
with causes that remain poorly understood but are commonly
attributed to aging and obesity. To gain insight into the eti-
ology of knee osteoarthritis, this study traces long-term trends
in the disease in the United States using large skeletal samples
spanning from prehistoric times to the present. We show that
knee osteoarthritis long existed at low frequencies, but since
the mid-20th century, the disease has doubled in prevalence.
Our analyses contradict the view that the recent surge in knee
osteoarthritis occurred simply because people live longer and
are more commonly obese. Instead, our results highlight the
need to study additional, likely preventable risk factors that
have become ubiquitous within the last half-century.
Author contributions: I.J.W., S.W., D.T.F., and D.E.L. designed
research; I.J.W., R.D.J.,
K.T.W., H.M., and R.J.W. performed research; I.J.W. and S.W.
analyzed data; and I.J.W.,
S.W., D.T.F., and D.E.L. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission. O.P. is a guest editor
invited by the Editorial
Board.
Freely available online through the PNAS open access option.
1To whom correspondence should be addressed. Email:
[email protected]
This article contains supporting information online at
www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1703856114/-/DCSupplemental.

9332–9336 | PNAS | August 29, 2017 | vol. 114 | no. 35
www.pnas.org/cgi/doi/10.1073/pnas.1703856114
http://crossmark.crossref.org/dialog/?doi=10.1073/pnas.1703856
114&domain=pdf
mailto:[email protected]
http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.170385611
4/-/DCSupplemental
4/-/DCSupplemental
Jennifer
Highlight
Jennifer
Sticky Note
(highlight in yellow) Most original research articles (primary
sources in science) provide a full title, list of authors, and the
author(s)affiliations (i.e. what research centres, organizations,
and/or universities/colleges they belong to) at the start of the
report.
Jennifer
Highlight
Jennifer
Sticky Note
(highlight in blue)
Research articles tend to be quite long and detailed; as a result,
it would be difficult to determine whether the research is
useful/interesting to the reader without reading the entire
report. So, the 'abstract' (sometimes labeled as such, other times
not) provides a short summary of the research objective(s),
methods, results, and conclusions. Reading the abstract (in this

article, it's the first 'paragraph' in bold fond) is a good place to
start because it gives an overview of what the report is going to
discuss.
Jennifer
Highlight
Jennifer
Sticky Note
(highlight in purple)
The first major section of a research report is the 'Introduction'
(also known as background, literature review, and possibly
other names). The introduction typically gives background
information (with reference to previous research) necessary to
understand the objectives and hypotheses of the study being
described. Most introductions start from very general
knowledge and move to more specific content, ending with a
descriptive statement of the study's objectives, and hypotheses.
In this particular report, the last paragraph of the introduction
describes the objectives/procedure ("Here, we investigate long-
term trends....").
Reading the introduction can be useful because researchers
often define important concepts or terms that they will use
throughout their report. For the purpose of our assignment, you
might just skim the introduction to provide some context for the
next section ("methods").
BMI on levels of the disease by comparing the prevalence of
knee OA among people who lived during the early industrial era
(19th to early 20th centuries) with that of people from the
modern postindustrial era (late 20th to early 21st centuries). We
studied knee OA in the largest available collections of cadaver-
derived skeletal remains of people of documented age, BMI,
sex,

and ethnicity. To further consider knee OA levels from an evo-
lutionary perspective, we also analyzed knee OA in a large
sample of archeological skeletons of prehistoric Native Ameri-
can hunter-gatherers (6000–300 B.P.) and early farmers (900–
300 B.P.). Although BMI is undocumented for prehistoric skel-
etons, the age at death and sex can be estimated, allowing us to
assess the prevalence of knee OA among older individuals in
these populations. The skeletal collections used in this study
are,
by necessity, samples composed of individuals who could not be
randomly selected and for whom we lack comprehensive de-
mographic and contextual information. Despite these
limitations,
these samples constitute the best available evidence for knee
OA
levels in the United States during earlier time periods to test if
prevalence of the disease is higher today than in the past.
Materials and Methods
Study Samples. The early industrial and postindustrial samples
studied in-
cluded complete skeletons of people aged 50 y and older who
lived in major
urban areas in the United States (Table S1). All individuals
were documented
as being of either European-American or African-American
ancestry. Early
industrial individuals (n = 1,581) were inhabitants of Cleveland,
Ohio and
St. Louis, Missouri who died between 1905 and 1940. BMI at
death was
recorded for 84% of these skeletons (n = 1,334). Postindustrial
individuals
(n = 819) lived in Albuquerque, New Mexico and Knoxville,
Tennessee and
died between 1976 and 2015. BMI at death was recorded for

64% of these
skeletons (n = 525). All cadavers were acquired by academic
institutions for
the purposes of medical and anatomical education and research.
Early in-
dustrial cadavers were of individuals whose bodies were
unclaimed at local
morgues or became property of the state; postindustrial
cadavers were
gathered through body donation programs. Occupation was
documented
for only 23% of individuals (n = 544), but records indicate that
differences
between samples reflect shifts in the US workforce between
early industrial
and postindustrial times, with the early industrial sample
comprising pri-
marily highly physically active laborers and the postindustrial
sample in-
cluding more service sector workers with less physically
demanding jobs (SI
Text). Cause of death was documented for 80% of individuals (n
= 1,918),
and differences between samples evince the epidemiological
transition be-
tween early industrial and postindustrial times, with most deaths
among
early industrial individuals caused by infectious diseases, such
as pneu-
monia and tuberculosis, whereas most deaths among
postindustrial indi-
viduals were caused by noninfectious diseases, such as cancer
and
atherosclerotic heart disease (SI Text). Skeletons with knee
joint articular
surfaces that were severely damaged postmortem were excluded

from the
study as were individuals with lower limb amputations.
The prehistoric sample included skeletons from eight
archeological sites (in
Alaska, California, New Mexico, Kentucky, and Ohio) of people
estimated to
be aged 50 y and older who were hunter-gatherers (n = 116) and
early
farmers (n = 60) (Table S1). Only skeletons sufficiently
preserved to examine
both the right and left knees were included. Sex assignment was
based on
dimorphic characteristics of the pubis that have been shown to
be 96%
accurate (18). Individuals were estimated to be ≥50 y old based
on age-
related changes in the configuration of the auricular surface of
the ilium
(19). This method has been shown to correctly exclude
individuals
younger than 50 y of age with 100% accuracy (20).
Unfortunately, esti-
mating age precisely beyond 50 y is not possible with available
osteo-
logical methods, and it is not possible to estimate accurately
BMI at
death from skeletal remains.
Knee OA Diagnosis. Diagnosis of knee OA was based on visual
identification of
the presence of eburnation on the articular surfaces of the right
or left distal
femur, proximal tibia, or patella. Eburnation is a sclerotic,
ivory-like reaction
of subchondral bone that occurs from bone-on-bone contact at

sites exposed
by advanced cartilage erosion (12, 21). In pathology studies of
skeletal re-
mains, eburnation is considered pathognomonic for moderate to
severe OA
(12, 22–24). Although it was not possible to assess knee OA
blinded to the
skeleton’s collection of origin, eburnation can be identified
with negligible
interobserver variation (SI Text). To avoid false-positive
diagnoses, individuals
exhibiting knee eburnation but also osteological signs of non-
OA arthritides,
such as rheumatoid arthritis, calcium pyrophosphate deposition
disease, and
spondyloarthropathy, were excluded. Osteophytes, bone spurs
that often
form at the margins of osteoarthritic joints (22–25), were
generally large and
expansive on eburnated knees but were not used as a diagnostic
criterion
for knee OA because interobserver variation in identifying
osteophytes in
skeletal samples is high (26), and they can lead to false-positive
diagnoses
(12); also, arthroplasty prostheses were not used to diagnose the
disease among
postindustrial individuals. Knee OA prevalence estimates for
our samples are
therefore underestimates of total disease prevalence, because
they do not
include mild (or early) cases of knee OA [e.g., cases that would
be classified
as two on the Kellgren–Lawrence scale (25)] or cases of the
disease where

arthroplasty prostheses obscure underlying eburnation.
Statistical Analyses. Log-binomial generalized linear models
(GLMs) were
used to estimate adjusted prevalence and prevalence ratios for
knee OA,
which are reported with 95% confidence intervals (95% CIs).
Prevalence is a
measure of effect size that varies as a function of the values of
predictors.
Here, we predict prevalence over a range of values of the
predictor of interest
while holding all other covariates constant at the sample mean.
A prevalence
ratio is a measure of effect size that is constant over the range
of the predictor
of interest while controlling for other covariates. Since
prevalence ratios are
multiplicative, they denote a rate of change (percentage change)
of the
response per unit increase in the predictor of interest. Model
goodness of fit
was assessed using the Hosmer–Lemeshow χ2 test, with
significance of in-
dividual estimates determined through two-sided Wald tests
with an alpha
level of 0.05 (Table S2).
Three separate GLMs were performed with a binary response
variable
indicating presence or absence of knee OA for each individual
but including
different explanatory variables. The first analysis included the
prehistoric,
early industrial, and postindustrial samples and controlled only
for sex effects,

since age and BMI were undocumented for prehistoric
individuals. The second
and third analyses included only the early industrial and
postindustrial
samples and additionally controlled for age, BMI, and ethnicity.
The second
analysis used all available individuals weighted equally,
whereas the third
analysis incorporated a subset of individuals who were
differentially
weighted based on optimal matching of covariate values
between the early
industrial and postindustrial samples. The analysis of matched
data was
performed as a sensitivity check to assess whether inferences
were robust to
sampling bias between the early industrial and postindustrial
samples. This
bias was evidenced by the large differences in average covariate
values
between these two samples in the unmatched data (Table 1). The
purpose of
matching is to approximate an experimental template, where the
matching
procedure mimics blocking before random group assignment to
balance
average covariate values between “target” and “comparator”
groups.
Separation of the estimation procedure into two steps simulates
the re-
search design of an experiment where no information on
outcomes is
known at the point of experimental design and randomization.
The non-
parametric matching procedure is therefore a data preprocessing
step that

replicates a randomized experiment with respect to observed
covariates
(27). Preprocessing was achieved by matching individuals from
the early
industrial and postindustrial samples that had a similar
propensity to be
included in the postindustrial sample based on covariate values
(SI Text).
Pruning nonmatches increased similarity in average covariate
values be-
tween the early industrial and postindustrial samples (Table 1)
and reduced
model dependency and bias (28). Weights for each individual
were con-
structed to estimate the average effect of interest on the
postindustrial
sample, with the early industrial sample weighted to look like
the post-
industrial sample. Analyses were conducted using R, version
3.3.2 (29).
Results
Long-Term Change in Knee OA Prevalence. Across all
individuals
analyzed (n = 2,576), the prevalence of knee OA was markedly
higher among individuals from the postindustrial era compared
with individuals from early industrial and prehistoric times,
with
females more affected than males (Fig. 1 A and C). After con-
trolling for sex, knee OA prevalence in the postindustrial
sample
(16%; 95% CI, 14–19%) was 2.6 times higher (95% CI, 2.1–3.4;
P < 0.001) than in the early industrial sample (6%; 95% CI, 5–
7%) and 2 times higher (95% CI, 1.3–3.3; P = 0.003) than in the
prehistoric sample (8%; 95% CI, 5–13%). Among postindustrial
individuals with knee OA, 42% (64/151) had the disease in both

knees, a 2.5-fold higher proportion (Fisher’s exact test: P =
0.042)
Wallace et al. PNAS | August 29, 2017 | vol. 114 | no. 35 | 9333
A
N
TH
R
O
P
O
LO
G
Y
4/-
/DCSupplemental/pnas.201703856SI.pdf?targetid=nameddest=S
T1
4/-
TXT
4/-
TXT
4/-
TXT
4/-

T1
4/-
TXT
4/-
T2
4/-
TXT
Jennifer
Highlight
Jennifer
Sticky Note
(highlighted in green)
The next major section of a research report is the 'Methods'
(sometimes 'Methods and materials', 'procedure', etc.).
Depending on the publication, the methods section may come
right after the introduction (as it does in this example), or it
may be placed later in the report (e.g. towards the end). The
methods section provides sufficient detail about the sampling
procedure, study design, and statistical analysis/data
manipulation that a reader could redo the study with a new set
of data.
For the purpose of our assignment, try not to get too bogged
down in descriptions of statistical analysis and very discipline-
specific methodology. Keep your focus on the level of
detail/understanding we have worked with in our course. The
report might not use the exact vocabulary you've seen in the
course (they might, though!), but the descriptions they use

should be recognizable.
Jennifer
Highlight
Jennifer
Sticky Note
(highlighted in red)
The third major section is the 'Results' section. As the heading
suggests, it is a description of the results. This section tends to
be pretty painful to read, especially if the researchers used
complex statistical analyses. But, it does present the results of
all comparisons, hypothesis tests, descriptive summaries, etc.
that they include in their report. You'll likely recognize means
and standard deviations, you might also see reference to
confidence intervals (e.g. 95% CI, 14-19) and hypothesis tests
(e.g. P=0.003). You'll also see tables and figures summarizing
key results to address research questions.
For the purpose of our assignment, it's best to read through the
Results section, looking for text that describes comparisons
useful to our assignment, and to explore the data
provided/summarized in tables and figures. Remember not to
get hung up on the details of the statistical analyses reported.
of bilateral cases of knee OA than among the diseased
individuals
in the prehistoric sample (17%; 3/18) and 1.4-fold higher
(Fisher’s
exact test: P = 0.058) compared with the early industrial sample
(30%; 28/94).
Temporal Change in Knee OA Prevalence Controlling for Age
and BMI.
To test whether the higher levels of knee OA in the post-
industrial era are attributable to greater longevity and higher

BMIs, we analyzed the subset of individuals in our samples for
Table 1. Sample composition and covariate balance before and
after matching
Variable
Unmatched analysis Matched analysis
Improvement, %Early industrial Postindustrial Early industrial
Postindustrial
Female/male ratio 0.17 0.41 0.46 0.39 71.3
Age, y 62.3 ± 9.7* 68.5 ± 10.4* 68.5 ± 9.8* 68.6 ± 10.4* 99.3
BMI, kg/m2 18.7 ± 4.2* 26.4 ± 8.0* 25.3 ± 6.8* 25.3 ± 6.3*
99.7
Ethnicity ratio† 0.32 0.023 0.039 0.024 95.0
Distance‡ 0.15 0.61 0.59 0.59 100
n 1334 525 857§ 500 100
*Mean ± SD.
†African American/European American ratio.
‡Distance measure is the propensity score of being in the
postindustrial sample, calculated using all observed
covariates.
§The 857 early industrial observations were down-weighted in
log-binomial models to equal the 500
observations from the postindustrial sample, thus giving an
effective sample size of 1,000 observations for
the analysis of matched data.
Female / male
Early industrial / prehistoric
Postindustrial / prehistoric

Postindustrial / early industrial
Knee OA prevalence ratio (%)
BMI
Age
Female / male
Postindustrial / early industrial
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Knee OA prevalence ratio
C
D
0
5
10
15
20
Pre
his
tor
ic

Ea
rly
in
du
str
ial
Po
stin
du
str
ial
K
n
e
e
O
A
p
re
va
le
n
ce

(
%
)
Ea
rly
in
du
str
ial
Po
stin
du
str
ial
A B
Fig. 1. Knee OA prevalence during different time periods. (A
and B) Knee OA prevalence from regression models controlling
for sex (A) as well as age, BMI,
sex, and ethnicity (B). Dark and light gray bars are from
unmatched and matched analyses, respectively (B). (C and D)
Knee OA prevalence ratios from re-
gression models including sex (C) as well as age, BMI, sex, and
ethnicity (D) as predictor variables. Black and light gray dots
are from unmatched and matched
analyses, respectively (D). Age and BMI were entered into
models as continuous variables, but effects are reported for 10-y
and 5-U intervals, respectively (D).

Whiskers represent 95% CIs. Ethnicity effects are reported in
Table S3.
9334 | www.pnas.org/cgi/doi/10.1073/pnas.1703856114 Wallace
et al.
4/-
T3
whom age and BMI were documented (n = 1,859). Individuals
from the postindustrial group were, on average, 6 y older and
had 41% higher BMIs than their early industrial counterparts
(Welch’s t test: P < 0.001 for both the age and BMI
comparisons)
(Table 1). Only 1% (13/1,334) of early industrial individuals
were
obese (BMI ≥ 30) and 6% (74/1,334) were overweight (25 ≤
BMI < 30) compared with 25% (132/525) and 24% (126/525) of
postindustrial individuals who were obese and overweight, re-
spectively (Fisher’s exact test: P < 0.001 for both the obese and
overweight comparisons). Nevertheless, in a model controlling
for
age, BMI, and other variables, knee OA prevalence in the post-
industrial sample (11%; 95% CI, 8–14%) remained 2.1 times
higher
(95% CI, 1.5–3.1; P < 0.001) than in the early industrial sample
(5%; 95% CI, 4–7%) (Fig. 1 B and D). Age and BMI were posi-
tively associated with knee OA prevalence (P < 0.001 for both
variables) (Fig. 1D), but at all ages, knee OA prevalence was at
least
twice as high in the postindustrial sample than in the early
industrial

sample, even after controlling for BMI (Fig. 2).
Temporal Change in Knee OA Prevalence Assessed Using
Matched
Samples. Matching individuals from the early industrial and
postindustrial samples by propensity score increased covariate
balance by 99% for age, 100% for BMI, 71% for sex, and 95%
for ethnicity (Table 1). In a model using these matched samples
and additionally controlling for age, BMI, and other variables,
knee OA prevalence in the postindustrial sample (15%; 95% CI,
12–19%) remained approximately twice as high (prevalence ra-
tio: 1.9; 95% CI, 1.1–3.5; P < 0.029) compared with the early
industrial sample (8%; 95% CI, 5–13%) (Fig. 1 B and D).
Discussion
To gain insight into the current high prevalence of OA in the
United States and other developed nations, this study examined
long-term trends in knee OA levels in the United States from
prehistoric times through the early industrial era to the modern
postindustrial era. These data show that knee OA long existed at
low frequencies, but since the mid-20th century, knee OA has
ap-
proximately doubled in prevalence, even after accounting for
the
effects of age and BMI. Our analyses therefore indicate that, al-
though knee OA prevalence has increased over time, today’s
high
levels of the disease are not, as commonly assumed, simply an
in-
evitable consequence of people living longer and more often
hav-
ing a high BMI. Instead, our analyses indicate the presence of
additional independent risk factors that seem to be either unique
to
or amplified in the postindustrial era.

Retrospective studies cannot directly test causation, but the
dramatic increase in knee OA prevalence in recent times raises
the
question of what these additional risk factors might be. Alleles
of
genes, such as GDF5, have been shown to influence knee OA
susceptibility (30), but the approximate doubling of knee OA
prevalence in just the last few generations proves that recent
envi-
ronmental changes have played a principal role. The results of
this
study are thus clinically significant because they indicate that
knee
OA may be more preventable than is currently supposed. Given
evidence that nearly all knee OA is associated with loading-
induced
damage to joint tissues (4), either because the loads are
abnormal or
the tissues are structurally weak, one especially important
source of
environmental change that warrants greater attention is whether
and how joint loading has altered. Trauma has presumably
always
predisposed some individuals to knee OA (8), as suggested by
the
predominance of unilateral knee OA since prehistoric times
(31),
and while joint overloading from high BMI has become common
only recently, our results indicate that the majority of knee OA
today is not caused by high BMI per se. Although altered loads
generated by walking more frequently on hard pavements (32)
or
with certain forms of footwear (33) might be factors, another
pos-
sibility that merits more study is physical inactivity, which has

be-
come epidemic during the postindustrial era. Less physically
active
individuals who load their joints less develop thinner cartilage
with
lower proteoglycan content (34, 35) as well as weaker muscles
re-
sponsible for protecting joints by stabilizing them and limiting
joint
reaction forces (36). Chronic low-grade inflammation, which is
ex-
acerbated by physical inactivity (37), modern diets rich in
highly
refined carbohydrates (38), and excessive adiposity (10), can
further
magnify and accelerate loading-induced damage to joint tissues
and
may also directly affect knee OA pathogenesis (5). Evaluating
which
of these or additional features of modern environments are re-
sponsible for today’s high knee OA levels is necessary.
This study has important limitations that need to be considered.
First, the samples analyzed, although large for their kind, were
constrained by the availability of well-curated skeletal
collections in
the United States, and it is plausible that these collections
exhibit
levels of knee OA that differ from the actual US population
prev-
alence. Second, BMI recorded at death is likely to
underestimate
average lifetime BMI, especially for individuals whose cause of
death was associated with somatic wasting. While discrepancy
be-
tween lifetime and postmortem BMI introduces error into the re-

lationship between BMI and knee OA, such error is likely to
have
been systematic rather than specific to a particular time period.
Third, although it is reasonable to infer that the postindustrial
in-
dividuals studied here were, on average, less physically active
and
consumed more proinflammatory diets than those from earlier
periods (39), direct data on these and other potential risk factors
are
not available for the individuals studied. Fourth, although
socio-
economic status was undocumented for individuals in this study,
the early industrial group likely included more relatively low-
income individuals than the postindustrial group. This differ-
ence, however, partly reflects important sociodemographic
shifts that occurred across the epidemiological transition be-
tween time periods (40). Fifth, BMI was unknown for pre-
historic individuals, and although sex is reliably determined,
age
estimates beyond 50 y old are imprecise. Thus, the prehistoric
samples could not be included in regression models that used
age and BMI as predictor variables, and although the modal age
of
adult death in living hunter-gatherers is 68–78 y old (41), we
cannot reject the hypothesis that knee OA levels are lower
among
prehistoric individuals than among postindustrial individuals,
partly because prehistoric individuals were, on average, younger
or
had lower BMIs.
Although the causes of OA in general and knee OA in partic-
ular are still not fully understood, the most important
conclusion

Earl
y ind
ustr
ial
Po
sti
nd
us
tria
l
0
5
10
15
20
25
50 55 60 65 70 75 80
Age (Years)
K
n
e
e

O
A
p
re
va
le
n
ce
(
%
)
Fig. 2. Age-related change in knee OA prevalence controlling
for BMI, sex,
and ethnicity. Shading represents 95% CIs.
Wallace et al. PNAS | August 29, 2017 | vol. 114 | no. 35 | 9335
A
N
TH
R
O
P
O
LO
G
Y

Jennifer
Highlight
Jennifer
Sticky Note
(highlighted in light blue)
The last major section is the 'Discussion'; this is typically the
most interesting section when reading a research report because
it discusses the results presented. That is, it places the results in
the context of the research questions, discusses their
similarities/differences to previous research results, and
attempts to explain why the results occurred. Discussion
sections also tend to allocate some time to discussing
limitations of their research study, and/or next steps/further
analyses that could be explored for greater understanding.
For the purpose of our assignment, reading the discussion might
be helpful in attempting to understand the results/comparisons
made, as well as any concerns/limitations the researchers
identified with their data.
of this study is that the recent increase in knee OA levels cannot
simply be considered an inevitable consequence of people living
longer, but instead is the result of modifiable risk factors, in-
cluding but not limited to high BMI, that have become more
common since the mid-20th century. From an evolutionary
perspective, knee OA thus fits the criteria of a “mismatch dis-
ease” that is more prevalent or severe because our bodies are
inadequately or imperfectly adapted to modern environments
(39). Intriguingly, other well-studied mismatch diseases, such
as
hypertension, atherosclerotic heart disease, and type 2 diabetes
(39), that also have become epidemic during the last few

decades
are strongly associated with knee OA (42), suggesting com-
mon causes and risk factors. Susceptibility to knee OA and
other
mismatch diseases is undoubtedly influenced by intrinsic
factors,
including age, sex, and genes, but the historical and
evolutionary
perspective afforded by our data underscores that many modern
cases of knee OA may be preventable. Prevention, however, will
require a reappraisal of potential risk factors that have emerged
or intensified only very recently. As with other mismatch dis-
eases, it is likely that any effective prevention strategy will in-
volve adjusting physical activity patterns and diets to
approximate
more closely the lifestyle conditions under which our species
evolved.
ACKNOWLEDGMENTS. We thank the curatorial staffs of
institutions housing
the skeletal collections analyzed, including the American
Museum of Natural
History, the Cleveland Museum of Natural History, the
Department of
Anthropology at San Jose State University, the Forensic
Anthropology
Center at the University of Tennessee, the Maxwell Museum of
Anthropol-
ogy at the University of New Mexico, the National Museum of
Natural
History, the Peabody Museum at Harvard University, and the W.
S. Webb
Museum of Anthropology at the University of Kentucky. We
also thank
Michèle Morgan for providing age estimates for the prehistoric

skeletons
from New Mexico, and Ashley Brennaman for providing data
used to assess
interobserver agreement for eburnation identification. This work
was sup-
ported by the Hintze Family Charitable Foundation and the
American School
of Prehistoric Research (Harvard University).
1. Murray CJ, et al. (2013) The state of US health, 1990-2010:
Burden of diseases, injuries,
and risk factors. JAMA 310:591–608.
2. Vos T, et al. (2012) Years lived with disability (YLDs) for
1160 sequelae of 289 diseases
and injuries 1990-2010: A systematic analysis for the Global
Burden of Disease Study
2010. Lancet 380:2163–2196.
3. Lawrence RC, et al. (2008) Estimates of the prevalence of
arthritis and other rheumatic con-
ditions in the United States. Part II. Arthritis Rheum 58:26–35.
4. Felson DT (2013) Osteoarthritis as a disease of mechanics.
Osteoarthritis Cartilage 21:10–15.
5. Robinson WH, et al. (2016) Low-grade inflammation as a key
mediator of the path-
ogenesis of osteoarthritis. Nat Rev Rheumatol 12:580–592.
6. Bijlsma JWJ, Berenbaum F, Lafeber FPJG (2011)
Osteoarthritis: An update with rele-
vance for clinical practice. Lancet 377:2115–2126.
7. Nguyen US, et al. (2011) Increasing prevalence of knee pain
and symptomatic knee

osteoarthritis: Survey and cohort data. Ann Intern Med
155:725–732.
8. Felson DT, et al. (2000) Osteoarthritis: New insights. Part 1:
The disease and its risk
factors. Ann Intern Med 133:635–646.
9. Loeser RF, Collins JA, Diekman BO (2016) Ageing and the
pathogenesis of osteoar-
thritis. Nat Rev Rheumatol 12:412–420.
10. Wluka AE, Lombard CB, Cicuttini FM (2013) Tackling
obesity in knee osteoarthritis.
Nat Rev Rheumatol 9:225–235.
11. Cross M, et al. (2014) The global burden of hip and knee
osteoarthritis: Estimates from
the Global Burden of Disease 2010 study. Ann Rheum Dis
73:1323–1330.
12. Rogers J, Dieppe P (2003) Paleopathology of osteoarthritis.
Osteoarthritis, eds
Brandt KD, Doherty M, Lohmander LS (Oxford Univ Press,
Oxford), 2nd Ed, pp 57–65.
13. Larsen CS (1982) The anthropology of St. Catherines Island.
3. Prehistoric human bi-
ological adaptation. Anthropol Pap Am Mus Nat Hist 57:162–
270.
14. Waldron HA (1991) Prevalence and distribution of
osteoarthritis in a population from
Georgian and early Victorian London. Ann Rheum Dis 50:301–
307.
15. Rogers J, Dieppe P (1994) Is tibiofemoral osteoarthritis in
the knee joint a new dis-

ease? Ann Rheum Dis 53:612–613.
16. Webb S (1995) Palaeopathology of Aboriginal Australians:
Health and Disease Across
a Hunter-Gatherer Continent (Cambridge Univ Press,
Cambridge, UK).
17. Inoue K, et al. (2001) Prevalence of large-joint
osteoarthritis in Asian and Caucasian
skeletal populations. Rheumatology (Oxford) 40:70–73.
18. Phenice TW (1969) A newly developed visual method of
sexing the os pubis. Am J Phys
Anthropol 30:297–301.
19. Lovejoy CO, Meindl RS, Pryzbeck TR, Mensforth RP (1985)
Chronological meta-
morphosis of the auricular surface of the ilium: A new method
for the determination
of adult skeletal age at death. Am J Phys Anthropol 68:15–28.
20. Murray KA, Murray T (1991) A test of the auricular surface
aging technique. J Forensic
Sci 36:1162–1169.
21. Pritzker KP, et al. (2006) Osteoarthritis cartilage
histopathology: Grading and staging.
Osteoarthritis Cartilage 14:13–29.
22. Cockburn A, Duncan H, Riddle JM (1979) Arthritis, ancient
and modern: Guidelines for
field workers. Henry Ford Hosp Med J 27:74–79.
23. Jurmain R (1999) Stories from the Skeleton: Behavioral
Reconstruction in Human

Osteology (Gordon and Breach, Amsterdam).
24. Waldron T (2009) Paleopathology (Cambridge Univ Press,
Cambridge, UK).
25. Kellgren JH, Lawrence JS (1957) Radiological assessment
of osteo-arthrosis. Ann
Rheum Dis 16:494–502.
26. Waldron T, Rogers J (1991) Inter-observer variation in
coding osteoarthritis in human
skeletal remains. Int J Osteoarchaeol 1:49–56.
27. Morgan SL, Winship C (2015) Counterfactuals and Causal
Inference (Cambridge Univ
Press, Cambridge, UK), 2nd Ed.
28. Ho DE, Imai K, King G, Stuart EA (2007) Matching as
nonparametric preprocessing for
reducing model dependence in parametric causal inference. Polit
Anal 15:199–236.
29. R Core Team (2016) R: A Language and Environment for
Statistical Computing (R
Foundation for Statistical Computing, Vienna).
30. Valdes AM, et al. (2011) The GDF5 rs143383 polymorphism
is associated with osteoarthritis
of the knee with genome-wide statistical significance. Ann
Rheum Dis 70:873–875.
31. Davis MA, Ettinger WH, Neuhaus JM, Cho SA, Hauck WW
(1989) The association of
knee injury and obesity with unilateral and bilateral
osteoarthritis of the knee. Am J
Epidemiol 130:278–288.

32. Radin EL, Orr RB, Kelman JL, Paul IL, Rose RM (1982)
Effect of prolonged walking on
concrete on the knees of sheep. J Biomech 15:487–492.
33. Kerrigan DC, Todd MK, Riley PO (1998) Knee
osteoarthritis and high-heeled shoes.
Lancet 351:1399–1401.
34. Kiviranta I, Tammi M, Jurvelin J, Säämänen A-M, Helminen
HJ (1988) Moderate
running exercise augments glycosaminoglycans and thickness of
articular cartilage in
the knee joint of young beagle dogs. J Orthop Res 6:188–195.
35. Urquhart DM, et al. (2011) What is the effect of physical
activity on the knee joint? A
systematic review. Med Sci Sports Exerc 43:432–442.
36. Roos EM, Herzog W, Block JA, Bennell KL (2011) Muscle
weakness, afferent sensory
dysfunction and exercise in knee osteoarthritis. Nat Rev
Rheumatol 7:57–63.
37. Handschin C, Spiegelman BM (2008) The role of exercise
and PGC1α in inflammation
and chronic disease. Nature 454:463–469.
38. Giugliano D, Ceriello A, Esposito K (2006) The effects of
diet on inflammation: Em-
phasis on the metabolic syndrome. J Am Coll Cardiol 48:677–
685.
39. Lieberman DE (2013) The Story of the Human Body:
Evolution, Health, and Disease
(Pantheon, New York).

40. Floud R, Fogel RW, Harris B, Hong SC (2011) The
Changing Body: Health, Nutrition, and Human
Development in the Western World Since 1700 (Cambridge
Univ Press, Cambridge, UK).
41. Gurven M, Kaplan H (2007) Longevity among hunter-
gatherers: A cross-cultural ex-
amination. Popul Dev Rev 33:321–365.
42. Zhuo Q, Yang W, Chen J, Wang Y (2012) Metabolic
syndrome meets osteoarthritis.
Nat Rev Rheumatol 8:729–737.
43. Church TS, et al. (2011) Trends over 5 decades in U.S.
occupation-related physical
activity and their associations with obesity. PLoS One
6:e19657.
44. Tudor-Locke C, Ainsworth BE, Washington TL, Troiano R
(2011) Assigning metabolic equiva-
lent values to the 2002 census occupational classification
system. J Phys Act Health 8:581–586.
45. Omran AR (1971) The epidemiologic transition. A theory of
the epidemiology of
population change. Milbank Mem Fund Q 49:509–538.
46. Gaziano JM (2010) Fifth phase of the epidemiologic
transition: The age of obesity and
inactivity. JAMA 303:275–276.
47. National Center for Health Statistics (2000) Leading Causes
of Death, 1900-1998.
Available at
https://www.cdc.gov/nchs/data/dvs/lead1900_98.pdf. Accessed

June 15,
2017.
48. Jones DS, Podolsky SH, Greene JA (2012) The burden of
disease and the changing task
of medicine. N Engl J Med 366:2333–2338.
49. Brennaman AL (2014) Examination of Osteoarthritis for
Age-at-Death Estimation in a
Modern Population. MS thesis (Boston Univ School of
Medicine, Boston).
50. Cohen J (1960) A coefficient of agreement for nominal
scales. Educ Psychol Meas 20:37–46.
51. Landis JR, Koch GG (1977) The measurement of observer
agreement for categorical
data. Biometrics 33:159–174.
52. Gelman A, Hill J (2007) Data Analysis Using Regression
and Multilevel/Hierarchical
Models (Cambridge Univ Press, Cambridge, UK).
53. Austin PC, Stuart EA (September 1, 2015) Estimating the
effect of treatment on binary
outcomes using full matching on the propensity score. Stat
Methods Med Res.
54. Stuart EA (2010) Matching methods for causal inference: A
review and a look for-
ward. Stat Sci 25:1–21.
55. Ho DE, Imai K, King G, Stuart EA (2011) MatchIt:
Nonparametric preprocessing for
parametric causal inference. J Stat Softw 42:1–28.

9336 | www.pnas.org/cgi/doi/10.1073/pnas.1703856114 Wallace
et al.
https://www.cdc.gov/nchs/data/dvs/lead1900_98.pdf
Jennifer
Highlight
Jennifer
Sticky Note
(highlighted in gray)
At the end of the report, you'll find the literature cited
(References) section; this typically follows an
Acknowledgments section and possibly some other minor
sections. You would have noticed that the researchers reference
ideas, data, methodology, and arguments from other researchers
throughout the research report. The Literature Cited section
provides the content on those citations, in a format required by
the publication.
For our purposes, you need not review the information in the
Literature Cited unless you read something in the research
report that interests you and you want to look up more content
on it for your own purposes!
Stat 1023/2037A – FW19 Assignment 2
Assignment 2:
Clarifying Statistical Research
Introduction

Throughout this course, we discuss ways in which data and
statistical analysis could be manipulated or
reported in manners which may be deceptive, or in the least,
ambiguous. We focus on learning
foundational concepts in statistics to equip you with the
background necessary to ask the ‘right
questions’ before making decisions based on the results of
research encountered in your schooling, life,
and/or careers. One way to demonstrate your understanding of
statistical concepts is to interpret and
explain research characteristics in layperson’s language. This
assignment, therefore, is set up to evaluate
your ability to apply these skills.
Learning Objectives
• Demonstrate an understanding of statistical vocabulary and
concepts when presented in an
original research report;
• Experience the process of searching for, and reading the
original research report upon which a
media story is based;
• Accurately explain statistical concepts and background using
non-statistical language within a
novel context.
Assignment
Media presents the results of statistical research in short articles
that are read by the general public; we
refer to these articles as ‘media stories’ in our course. However,
these stories often summarize only key
results (with little content on context and/or sampling and study
design) to present a more sensational

story. As a consequence, relevant information and details—
necessary to make rational decisions based
on the research—are typically lacking. Your assignment is to
take one of the following media stories,
find the original published source, and rewrite the media story
(using language suitable for a layperson)
to make it more informative.
Steps to complete the assignment
1. Choose one of the following media stories; it’s a good idea to
read through all three options
before settling on which one you will work with. A keyword for
each media story has been
provided to help when you are uploading your Assignment to
Gradescope:
Option 1: altruism
“We’d rather harm ourselves than others, electric shock study
finds.”
https://www.tvnz.co.nz/one-news/world/wed-rather-harm-
ourselves-than-others-electric-
shock-study-finds
Option 2: pregnant
“Pregnant women with morning sickness are more likely to use
marijuana: Study”
https://www.huffingtonpost.ca/2018/08/22/marijuana-
pregnancy-
study_a_23507102/?utm_hp_ref=ca-living

ourselves-than-others-electric-shock-study-finds
ourselves-than-others-electric-shock-study-finds
pregnancy-study_a_23507102/?utm_hp_ref=ca-living
pregnancy-study_a_23507102/?utm_hp_ref=ca-living
admin
高亮
Option 3: sex
“Those good vibes you feel after sex can last for 48 hours,
study finds”
https://www.huffingtonpost.ca/entry/those-good-vibes-you-
feel-after-sex-can-last-for-48-
hours-study-finds_n_58d3f3b2e4b0b22b0d1aa255
2. Read the media story very carefully so you understand all the
information that it contains. As
you read, note any information about the source of the research
(e.g. researcher
names/affiliations, original publication source, keywords for the
research objective/variables,
etc.).
3. Think carefully about the information provided in the media
story and what is missing or
ambiguous, or what you still have questions about regarding the
study. That is, using what
you’ve learned in this course, try to identify what information

or details someone reading the
media story would need to know about the sampling/study
design and results to make informed
decisions based on the research.
4. Find the original research report describing the study (i.e. the
primary source or research
article); use the information about the source of the research
that you noted in step 2 to help in
this search. If you’re not sure you’ve found the original
research report, take a look at the
‘example of an original research report’ that was posted
alongside this instruction file. The
example also provides some descriptions of what’s included in a
research report and how to
approach reading one.
Be careful: Sometimes, a researcher has published several
research reports
about a particular topic. When you are looking for the original
research report for
this assignment, be sure to see if the report you find actually
matches the
information described in the media story. If you aren’t certain,
ask! It’s better to
double check before trying to write your assignment based on
the wrong report!
5. Read the original research report with the purpose of finding
the information/clarification you
identified as lacking in step 3. Make notes on this information
in your own words.

Note: research reports often use a lot of discipline-specific
jargon and complex
statistical analysis. Do NOT get bogged down or discouraged by
the language of
the research article. Stay focused on understanding the
information necessary to
explain what the researchers did and what were the results.
Again, if you need
help, ask!
6. Once you’ve done this background work, you can write your
assignment according to the
description/format detailed below.
Assignment Description
Write your own version of the media story, providing enough
detail/information about the study
design/results so that readers of your version would have the
necessary content to make informed
decisions about the study’s conclusions. In essence, you should
create a ‘fair’, unambiguous version of
the media story. However, you must write the
details/information in non-statistical jargon (i.e.
layperson’s terms) such that an individual with no background
in statistics would understand (you can
assume that the person has typical high school math knowledge,
if necessary). As well, keep in mind that
you are writing a media story (i.e. which tends to be more
conversational and sensational in tone), and
not simply summarizing the research. Consequently, you should
be creative (e.g. changing the title,
making up answers from interviews with the researchers, using
humour, etc.) provided your resulting

admin
高亮
admin
高亮
admin
高亮
admin
高亮
admin
高亮
admin
高亮
media story is accurate (i.e. correctly discusses the research and
findings…refer back to the learning
objectives for this assignment).
At the end of your assignment, include a citation to the original
research article on which the media
story is based. You may use any reference formatting format
you are familiar with, provided it gives full
information about the article. For example, the following would
be a suitable format:
Galperin, A., Haselton, M.G., Frederick, D.A., Poore, J., von
Hippel, W., Buss, D.M., and G.C. Gonzaga.

2012. Sexual regret: Evidence for evolved sex differences.
Archives of Sexual Behavior 42(7): 1145-1161.
Doi: 10.1007/s10508-012-0019-3
Your assignment must follow these formatting criteria (see the
Example assignment):
-and-a-half single-sided, 8.5” x 11”
pages. Any additional content may
not be graded.
etc.). Black font only, please.
(i.e. top, bottom, left, right)
complete sentences, using proper
English punctuation, spelling, and
grammar. Have someone who is not in our course edit it for you.
this, ask for help!
Marking Scheme
Your assignment will be graded based on the following criteria:
✓ Adherence to ‘Format’ criteria described above (4 marks);
✓ Evidence of understanding and accuracy of description of the
original research study design and
results (4 marks);
✓ Appropriate ‘voice’ for a media story (i.e. use of non-
statistical vocabulary that a non-statistics

student would understand; maintaining the tone of a media story
as opposed to a research
article or summary of research) (3 marks);
✓ Inclusion of original article citation at the end of your
assignment (2 marks);
Questions you should ask yourself before submitting:
study design, and results needed to
make the media story better/more informative? Have I missed
any relevant ideas/information
that would be necessary to make an informed opinion on the
research described in the media
story?
demonstrating I understand the study
design and results?
media story written in language that someone who has
not taken statistics would
understand?
a summary of research?
(remember, you are just writing a new version of the media
story that is more informative—it
still is meant to attract attention and/or be entertaining)
spelling, and had someone else not in
statistics edit it?

ave I included the citation for the original article at the end
of my assignment?
admin
高亮
admin
高亮
admin
高亮
admin
高亮
admin
高亮
admin
高亮
admin
高亮
Note Carefully: As you are working on your assignment, email
it to yourself occasionally, or save a
copy on a USB key or to a ‘Cloud’ (e.g. Dropbox). Don’t put
yourself in a situation where your
computer crashes and you lose your work right before the

deadline.
Writing in your own words
One of the major challenges of this assignment—or any
situation in which you must describe something
someone else has already done—is writing in your own words.
You will be working with two pieces of
writing from other individuals, (i) the original media story, and
(ii) the original research report, in order
to make your own piece of writing. In this type of situation, it
can be very easy to (accidentally or
otherwise) use the pieces of the other writing in your own,
especially if you take notes on the
story/report by copying sentences into your own file. To avoid
getting into this situation, I suggest the
following strategies:
• Writing something in your own words based on information
from other people’s words takes
some work and time—especially when you have to understand
the content of an original
research report written for discipline experts;
• Read the original media story several times so that you know
what it talks about without
looking at the actual page; this means you won’t really have to
look back at the media story
again for reference when reading the original research report, or
when writing your own
version of the media story (which will limit the opportunity to
use the media story’s phrasing);
• Make notes on the original research report in your own words.
It is SO tempting to just copy
relevant phrases into your notes as you read, with the purpose

of collecting the relevant
information for rewording later. But, as time passes you might
find yourself in two situations:
(i) you might forget what is your words versus copied from the
report, and (ii) you might be
pressed for time and struggle to rephrase your notes. If your
notes are always in your own
words, you’ve already done the hard work!
In the end, it’s always easier to write something in your own
words if you’ve worked hard to understand
the research you’ve been reading. If you understand it, you can
write freely from your memory, without
referring to any notes (whether those notes are your words, or
from the original sources).
Comment on Referencing
For the purpose of this assignment, you should not need to make
any references/citations within your
writing except the inclusion of the citation for the original
research article. The content of your
assignment will be based on the media story you chose (and
associated original research report) and the
textbook/lecture material presented in this course. You can
assume I know that the material you are
presenting is from these sources. You do have the freedom to
‘make up’ quotations from the
researchers; use standard quotation format (e.g. refer back to
how quotes from researchers are typically
formatted in the media story). You can assume I know that any
quotes you include are made up for the
purpose of making your media story more interesting/creative.
There should be no reason to quote directly from the original
research report OR the media story;

consequently, you should not be using any true quotes (beyond
what you have ‘made up’ to maintain a
media story tone) in your assignment.
How to submit this assignment:
Your assignment is submitted digitally (i.e. not on paper) in
TWO (2) places. You need to submit to BOTH
of these portals:
1. On OWL, through the “Assignment (OWL submissions)” tool.
2. On Gradescope. You were emailed an account set up email to
your UWO email address on
September 29. Follow that email to set up your Gradescope
account, and to submit your
assignment.
Academic Integrity and Plagiarism
You must complete this assignment on your own and in your
own words; no collaboration with peers is
permitted at any stage of the assignment. Your assignment will
be automatically submitted to Turnitin
(as per the course syllabus) for an originality report when you
submit your assignment through our OWL
website as described above (you do not have to and must not
separately submit your assignment to
Turnitin).
What does Turnitin do?

Turnitin compares the content of the assignment you submit
against websites, online databases and
repositories of past assignments submitted to Western as well as
other universities. It checks for ‘textual
similarity’ between your assignment and these
databases/repositories by matching similar phrases and
sentences. It then generates an ‘originality report’ for your
assignment, which indicates the percentage
of your assignment which has textual similarity to other
sources. An originality report of 100% indicates
your entire assignment was copied from other sources.
Why use Turnitin for Stat 1023/2037?
Turnitin is being used to ensure that students complete and
write their assignments in their own words.
This is relevant to Stat 1023/2037 as this assignment is part of
your course mark; the work you submit
should reflect YOUR understanding and ability.
How does my Turnitin ‘originality report’ affect my mark for
Stat 1023/2037?
If you have written your assignment in your own words, your
submission to Turnitin shouldn’t affect
your mark on the assignment at all!
In the event that the analysis provided by Turnitin indicates a
high degree of textual similarity, the
situation will be dealt with as described in the University’s
procedure for handling scholastic offences
(see the Western Academic Calendar).
Now I’m worried that I don’t know what plagiarism is or
whether I’m paraphrasing correctly. Help!
If you are concerned about whether you are paraphrasing
information correctly or adequately, or are
uncertain what writing practices constitute plagiarism, ask for

help! Just be sure to ask for this help with
plenty of time before the submission deadline!
Need help on the assignment?
• If you don’t fully understand what you need to do for this
assignment:
o post a question to the Forum under ‘Assignment 2’ (be sure to
use an informative title!);
o come to our course drop-in hours to ask for clarification
and/or help.
• If you aren’t sure what constitutes plagiarism, find out! There
are plenty of resources available:
o come to our course drop-in hours;
o speak with a librarian at the Research Help desk and/or review
the information on
plagiarism from the Western libraries
(http://www.lib.uwo.ca/tutorials/plagiarism);
o check out this website: http://www.plagiarism.org/plagiarism-
101/what-is-plagiarism/
• If you need help with spelling, grammar, punctuation, editing
a draft, clarity, style, or writing in
general, visit the Writing Support Centre
(http://www.sdc.uwo.ca/writing/). The people and
services available are a great resource for writers of all skill
levels. They have individual

appointments as well as drop-in hours (see the website for
information).
• If you don’t know how to set the preferences on your
computer software (e.g. Microsoft Word,
etc) so that the margins, line spacing, etc are correct, ask for
help during our course drop-in
hours or on the Forum;
• If you need help finding the original research report or
obtaining access to the report, Western’s
library team can help! Their research help site
www.lib.uwo.ca/services/research_help.html is a
good place to start; it even has videos on how to get full text
versions of original articles. You
might also have success walking in to one of the libraries and
asking for help at the research
desk. Alternatively, if you are struggling with finding
information about the original research
report in the first place, stop by our course drop-in hours for
help.
• If you need help writing a citation for the original research
report (i.e. identifying what the
components of a citation are), just ask! Post to the Forums, or
come to drop-in hours.
• If you have any other questions, just ask! Come to drop-in
hours, or use the Forums.
http://www.lib.uwo.ca/tutorials/plagiarism
http://www.plagiarism.org/plagiarism-101/what-is-plagiarism/

http://www.sdc.uwo.ca/writing/
http://www.lib.uwo.ca/services/research_help.html

Stat 1023 Assignment 2 Example Assignment 2 - comments .docx

Recommended

Recommended

More Related Content

Similar to Stat 1023 Assignment 2 Example Assignment 2 - comments .docx

Similar to Stat 1023 Assignment 2 Example Assignment 2 - comments .docx (7)

More from susanschei

More from susanschei (20)

Recently uploaded

Recently uploaded (20)

Stat 1023 Assignment 2 Example Assignment 2 - comments .docx