Abstract
Background on Project
We wanted to analyze student awareness of sustainability initiatives at Macalester College. While there are select student organizations working on sustainabilty intiatives on campus, we wanted to see how the student body approaches sustainability issues. We decided to look at a few different projects: (1) composting, (2) recycling, and (3) the EcoHouse. This was the first year that Macalester introduced composting and during orientation this past fall, the first-year students received training about the new composting infrastructure. For the composting initiative, we hypothesize that first years will have greater awareness than the upper classmen. For the other two initiatives, we hypothesize that awareness increases as students spend more time on Macalester's campus.
Survey Description
Our survey has twenty-one questions. There is only one yes or no question and the rest are multiple choice questions. Here is the link to our survey [(https://docs.google.com/a/macalester.edu/spreadsheet/viewform?fromEmail=true&formkey=dGp6UzdPRWxwQ2JKOG1BcWJWYTlkRmc6MQ)]. In addition to demographic questions including year, major and housing, we asked questions pertaining to degree of awareness of sustianability initiatives on campus.
myCSVLink = "https://docs.google.com/spreadsheet/pub?key=0Aksa37wkNYH6dGp6UzdPRWxwQ2JKOG1BcWJWYTlkRmc&output=csv"
d = fetchGoogle(myCSVLink)
The names of variables generated by Google Form are too verbose.
origNames = names(d)
origNames
## [1] "Timestamp"
## [2] "What.is.your.class.year."
## [3] "Are.you.an.international.student"
## [4] "How.often.do.you.recycle."
## [5] "How.often.do.you.compost."
## [6] "Did.you.know.that..Macalester.recently.committed.to.increase.its.purchasing.of.sustainable..local..fair.and.humane.food."
## [7] "Are.there.facilities.on.campus.for.students.to.compost."
## [8] "Are.there.facilities.on.campus.for.students.to.recycle."
## [9] "Does.Macalester.use.worm.compost.bins."
## [10] "Does.food.sourcing.affect.your.food.purchasing.decisions."
## [11] "Can.the.following.items.be.recycled....Paper."
## [12] "Can.the.following.items.be.recycled....Cardboard."
## [13] "Can.the.following.items.be.recycled....Plastic."
## [14] "Can.the.following.items.be.recycled....Aluminum."
## [15] "Can.the.following.items.be.recycled....Glass.."
## [16] "What.is.the.highest.level.type.of.plastics.that.can.be.recycled..SPI.Codes.."
## [17] "How.often.do.you.compost.outside.of.Macalester."
## [18] "Are.you.aware.of.Macalester.s.housing.option.committed.to.environmental.sustainability."
## [19] "Have.you.attended.an.event.sponsored.by.the.Eco.House."
## [20] "If.convenient.would.you.have.your.own.urban.farm..farm.plot."
## [21] "How.much.time.would.you.be.willing.to.dedicate.to.your.farm.farm.plot."
## [22] "What.is.your.academic.major."
## [23] "Do.you.have.a.second.major..if.so.please.select.from.the.list."
## [24] "Where.do.you.live."
Use this style to complete the change of names, e.g.
names(d)[2] = "year"
names(d)[3] = "international"
names(d)[4] = "freq.recycle"
names(d)[5] = "freq.compost"
names(d)[6] = "RFC.remove"
names(d)[7] = "fac.compost.remove"
names(d)[8] = "fac.recycle"
names(d)[9] = "worm.bins"
names(d)[10] = "source.l2.purchase"
names(d)[11] = "recycle.paper"
names(d)[12] = "recycle.cardboard"
names(d)[13] = "recycle.plastic"
names(d)[14] = "recycle.aluminum"
names(d)[15] = "recycle.glass"
names(d)[16] = "level.plastic"
names(d)[17] = "freq.compost.home"
names(d)[18] = "aware.ecohouse"
names(d)[19] = "event.ecohouse"
names(d)[20] = "own.farm"
names(d)[21] = "time.farm"
names(d)[22] = "major"
names(d)[23] = "second.major.remove"
names(d)[24] = "housing"
Checking our work:
paste(origNames, names(d), sep = "-->>")
## [1] "Timestamp-->>Timestamp"
## [2] "What.is.your.class.year.-->>year"
## [3] "Are.you.an.international.student-->>international"
## [4] "How.often.do.you.recycle.-->>freq.recycle"
## [5] "How.often.do.you.compost.-->>freq.compost"
## [6] "Did.you.know.that..Macalester.recently.committed.to.increase.its.purchasing.of.sustainable..local..fair.and.humane.food.-->>RFC.remove"
## [7] "Are.there.facilities.on.campus.for.students.to.compost.-->>fac.compost.remove"
## [8] "Are.there.facilities.on.campus.for.students.to.recycle.-->>fac.recycle"
## [9] "Does.Macalester.use.worm.compost.bins.-->>worm.bins"
## [10] "Does.food.sourcing.affect.your.food.purchasing.decisions.-->>source.l2.purchase"
## [11] "Can.the.following.items.be.recycled....Paper.-->>recycle.paper"
## [12] "Can.the.following.items.be.recycled....Cardboard.-->>recycle.cardboard"
## [13] "Can.the.following.items.be.recycled....Plastic.-->>recycle.plastic"
## [14] "Can.the.following.items.be.recycled....Aluminum.-->>recycle.aluminum"
## [15] "Can.the.following.items.be.recycled....Glass..-->>recycle.glass"
## [16] "What.is.the.highest.level.type.of.plastics.that.can.be.recycled..SPI.Codes..-->>level.plastic"
## [17] "How.often.do.you.compost.outside.of.Macalester.-->>freq.compost.home"
## [18] "Are.you.aware.of.Macalester.s.housing.option.committed.to.environmental.sustainability.-->>aware.ecohouse"
## [19] "Have.you.attended.an.event.sponsored.by.the.Eco.House.-->>event.ecohouse"
## [20] "If.convenient.would.you.have.your.own.urban.farm..farm.plot.-->>own.farm"
## [21] "How.much.time.would.you.be.willing.to.dedicate.to.your.farm.farm.plot.-->>time.farm"
## [22] "What.is.your.academic.major.-->>major"
## [23] "Do.you.have.a.second.major..if.so.please.select.from.the.list.-->>second.major.remove"
## [24] "Where.do.you.live.-->>housing"
require(plyr) # just need to do once, like require(mosaic)
levels(d$year)
## [1] "First-year" "Junior" "Senior" "Sophomore"
# check that the spelling matches in the next line
newLevels = c("First-year"="Fre","Sophomore"="Soph",
"Junior"="Jun","Senior"="Sen")
Now you will assign these new levels to your variable:
d$year = revalue(d$year, newLevels)
levels(d$year) = newLevels
Many of the survey questions are on a Likert Scale. You will want to simplify the names and also to tell R that there is a natural order. For example, the Web variable in our survey has a natural ordering.
Here is the renaming step:
likertLevels = c("Very aware"="Very" ,
"Somewhat aware"="Some",
"Not aware"="None")
d$aware.ecohouse = revalue(d$aware.ecohouse, likertLevels)
When you construct the translation (here called likertLevels), make sure to order it in the natural way, from one end to the other. Here, the order goes “None”, “Little”,“Some”,“Very”. You often will be able to use the same translation across multiple variables.
Now, tell R that the variable is ordered:
d$aware.ecohouse = factor(d$aware.ecohouse, ordered = TRUE, levels = likertLevels)
When you look at the variable, you will see that the level ordering is as specified:
head(d$aware.ecohouse)
## [1] Very Some Very Very Very Very
## Levels: Very < Some < None
as.integer(d$freq.recycle)
## [1] 2 2 2 1 1 2 1 3 2 3 1 3 3 1 1 2 1 3
as.integer(d$freq.compost)
## [1] 3 4 2 4 3 3 2 3 4 2 3 4 2 4 4 4 1 2
as.integer(d$aware.ecohouse)
## [1] 1 2 1 1 1 1 1 2 3 3 1 2 2 1 3 2 1 1
as.integer(d$housing)
## [1] 3 3 3 1 1 3 3 3 3 3 1 3 3 2 3 3 2 4
Describing Our Survey Data
We got 18 responses. The following a the distribution of some of the important questions:
tally(~year | freq.recycle, data = d)
## freq.recycle
## year Always Often Sometimes
## Fre 0.0000 0.3333 0.8000
## Soph 0.1429 0.3333 0.2000
## Jun 0.5714 0.0000 0.0000
## Sen 0.2857 0.3333 0.0000
## Total 1.0000 1.0000 1.0000
tally(~year | freq.compost, data = d)
## freq.compost
## year Always Never Often Sometimes
## Fre 0.0000 0.4000 0.4000 0.2857
## Soph 1.0000 0.4000 0.0000 0.1429
## Jun 0.0000 0.0000 0.4000 0.2857
## Sen 0.0000 0.2000 0.2000 0.2857
## Total 1.0000 1.0000 1.0000 1.0000
tally(~year | aware.ecohouse, data = d)
## aware.ecohouse
## year Very Some None
## Fre 0.1000 0.6000 0.6667
## Soph 0.3000 0.2000 0.0000
## Jun 0.4000 0.0000 0.0000
## Sen 0.2000 0.2000 0.3333
## Total 1.0000 1.0000 1.0000
tally(~year | housing, data = d)
## housing
## year Off campus, apartment Off campus, house On campus, dorms
## Fre 0.0000 0.0000 0.5000
## Soph 0.0000 0.5000 0.1667
## Jun 1.0000 0.5000 0.0000
## Sen 0.0000 0.0000 0.3333
## Total 1.0000 1.0000 1.0000
## housing
## year On campus, other
## Fre 0.0000
## Soph 1.0000
## Jun 0.0000
## Sen 0.0000
## Total 1.0000
mod1 = lm(as.integer(freq.compost) ~ year + housing, data = d)
anova(mod1)
## Analysis of Variance Table
##
## Response: as.integer(freq.compost)
## Df Sum Sq Mean Sq F value Pr(>F)
## year 3 3.50 1.167 1.36 0.30
## housing 3 3.08 1.028 1.20 0.35
## Residuals 11 9.42 0.856
mod2 = lm(as.integer(freq.recycle) ~ year + housing, data = d)
anova(mod2)
## Analysis of Variance Table
##
## Response: as.integer(freq.recycle)
## Df Sum Sq Mean Sq F value Pr(>F)
## year 3 7.44 2.481 11.70 0.00095 ***
## housing 3 2.00 0.667 3.14 0.06902 .
## Residuals 11 2.33 0.212
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
summary(mod2)
##
## Call:
## lm(formula = as.integer(freq.recycle) ~ year + housing, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.667 0.000 0.000 0.333 0.500
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.67e+00 7.98e-01 2.09 0.0607 .
## yearSoph -6.67e-01 3.76e-01 -1.77 0.1039
## yearJun -6.67e-01 7.52e-01 -0.89 0.3944
## yearSen -1.17e+00 2.97e-01 -3.92 0.0024 **
## housingOff campus, house -3.17e-16 5.32e-01 0.00 1.0000
## housingOn campus, dorms 1.00e+00 7.75e-01 1.29 0.2235
## housingOn campus, other 2.00e+00 8.41e-01 2.38 0.0366 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.461 on 11 degrees of freedom
## Multiple R-squared: 0.802, Adjusted R-squared: 0.694
## F-statistic: 7.42 on 6 and 11 DF, p-value: 0.0023
mod3 = lm(as.integer(aware.ecohouse) ~ year + housing, data = d)
anova(mod3)
## Analysis of Variance Table
##
## Response: as.integer(aware.ecohouse)
## Df Sum Sq Mean Sq F value Pr(>F)
## year 3 3.94 1.315 2.38 0.13
## housing 3 0.25 0.083 0.15 0.93
## Residuals 11 6.08 0.553
Analysis of Results -
Based on the regression report for model 1, we cannot reject the null hyposthesis that year has no effect of frequency of composting. All p-values are above .05. Our hypothesis that first year students are more likely to compost was not supported.
Based on the regression report for model 2, there is indication that our alternate hypothesis for recycling was not supported. Senior students were significantly less likely to recycle than other Macalester students.
We could not reject the null hypothesis for model 3. Year does not seem to be signficantly related to knowledge of the Ecohouse.