UQM paper figures
KZ
Feedback use
## Loading required package: bitopsMethods details
how many students in the cohort each semester
## BIOL1040 Sem1 BIOL1040 Sem2 BIOM2011 Sem1 BIOM2011 Sem2
## 733 972 251 92How many students looked at their reports?
How long do students leave their feedback open?
as mean +/- SEM in minutes
as mean +/- SEM in hours
## projects ls.mean ls.se
## 1 BIOL1040Sem1Report 1 3.9253858 0.010307411
## 2 BIOL1040Sem1Report 2 3.8234282 0.010809668
## 3 BIOL1040Sem1Report 3 3.7940786 0.010502834
## 4 BIOL1040Sem1Report 4 0.7151577 0.004828290
## 5 BIOL1040Sem2Report 1 3.2654889 0.006784010
## 6 BIOL1040Sem2Report 2 3.0790065 0.007191402
## 7 BIOL1040Sem2Report 3 0.4788550 0.003076460
## 8 BIOM2011Sem1Report 1 6.4423636 0.070072392
## 9 BIOM2011Sem1Report 2 1.3357410 0.024507403
## 10 BIOM2011Sem2Report 1 3.7142518 0.134024592
## 11 BIOM2011Sem2Report 2 0.5537070 0.030914024
open duration as histograms
NB log scale, so
## -4 -3 -2 -1 0 1 2 3
## "1 sec" "3 sec" "8 sec" "22 sec" "1 min" "3 min" "7 min" "20 min"
## 4 5 6 7 8
## "55 min" "2 hr" "7 hr" "18 hr" "50 hr"
Binning students based on open duration
bins at >1min, 1min-1hr, <1hr
## SubmissionID OpenDurationTotal.min project course sem
## 1 1486 56.23980 BIOL1040Sem1Report 1 BIOL1040 Sem1
## 2 1487 11.72763 BIOL1040Sem1Report 1 BIOL1040 Sem1
## 3 1488 59.98822 BIOL1040Sem1Report 1 BIOL1040 Sem1
## 4 1489 468.86465 BIOL1040Sem1Report 1 BIOL1040 Sem1
## 5 1490 1399.60237 BIOL1040Sem1Report 1 BIOL1040 Sem1
## report open.bin
## 1 Report 1 (1,60]
## 2 Report 1 (1,60]
## 3 Report 1 (1,60]
## 4 Report 1 (60,4.6e+03]
## 5 Report 1 (60,4.6e+03]##
## short medium long
## 559 2802 1920## StudentID SubmissionID MarkerID Publish.Time Final.Grade
## 1 S8110327 1711 T18 21/03/13 11:02 71
## 2 S8112239 2271 T25 27/03/13 13:30 79
## 3 S8233701 1755 T15 18/03/13 15:55 79
## 4 S8234819 2261 T03 25/03/13 5:41 75
## 5 S8236749 1929 T09 22/03/13 12:53 83
## project course sem report
## 1 BIOL1040Sem1Report 1 BIOL1040 Sem1 Report 1
## 2 BIOL1040Sem1Report 1 BIOL1040 Sem1 Report 1
## 3 BIOL1040Sem1Report 1 BIOL1040 Sem1 Report 1
## 4 BIOL1040Sem1Report 1 BIOL1040 Sem1 Report 1
## 5 BIOL1040Sem1Report 1 BIOL1040 Sem1 Report 1## SubmissionID StudentID MarkerID Publish.Time Final.Grade
## 1 10312 S8434173 T17 15/10/13 17:26 92
## 2 10314 S8646403 T01 5/10/13 13:01 65
## 3 10315 S8626753 T05 5/10/13 13:01 75
## 4 10316 S8633583 T06 18/10/13 8:08 54
## 5 10317 S8636705 T05 5/10/13 13:01 79
## project course sem report OpenDurationTotal.min
## 1 BIOL1040Sem2Report 3 BIOL1040 Sem2 Report 3 NA
## 2 BIOL1040Sem2Report 3 BIOL1040 Sem2 Report 3 NA
## 3 BIOL1040Sem2Report 3 BIOL1040 Sem2 Report 3 1.905667
## 4 BIOL1040Sem2Report 3 BIOL1040 Sem2 Report 3 NA
## 5 BIOL1040Sem2Report 3 BIOL1040 Sem2 Report 3 1.695567
## open.bin
## 1 <NA>
## 2 <NA>
## 3 medium
## 4 <NA>
## 5 medium## short medium long NA's
## 559 2802 1920 786Final grade by open duration
## means errors
## long 77.74407 0.2882501
## medium 78.33212 0.2635101
## short 78.31740 0.5792307
## unopened 67.26081 0.5872290Open duration by report
is there a sig diff in the amount of time students leave their open?
## [1] "Level 2 - semesters pooled"##
## Welch Two Sample t-test
##
## data: df[, 10] by df[, 9]
## t = 6.0832, df = 266.001, p-value = 4.084e-09
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 184.8511 361.7757
## sample estimates:
## mean in group Report 1 mean in group Report 2
## 342.24975 68.93636## [1] "Level 1 - semesters pooled"## Df Sum Sq Mean Sq F value Pr(>F)
## df[, 9] 3 17120573 5706858 51.02 <2e-16 ***
## Residuals 4738 529944612 111850
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 574 observations deleted due to missingness## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = df[, 10] ~ df[, 9])
##
## $`df[, 9]`
## diff lwr upr p adj
## Report 2-Report 1 -12.60544 -44.86734 19.65647 0.7469987
## Report 3-Report 1 -94.64860 -127.41763 -61.87956 0.0000000
## Report 4-Report 1 -175.74088 -218.50975 -132.97202 0.0000000
## Report 3-Report 2 -82.04316 -114.77301 -49.31331 0.0000000
## Report 4-Report 2 -163.13544 -205.87430 -120.39659 0.0000000
## Report 4-Report 3 -81.09228 -124.21523 -37.96934 0.0000083## [1] "Level 1 Semester 1"## Df Sum Sq Mean Sq F value Pr(>F)
## df[, 9] 3 16203646 5401215 42 <2e-16 ***
## Residuals 2390 307324667 128588
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 267 observations deleted due to missingness## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = df[, 10] ~ df[, 9])
##
## $`df[, 9]`
## diff lwr upr p adj
## Report 2-Report 1 -17.9152000 -71.17126 35.34086 0.8230058
## Report 3-Report 1 -18.6585637 -72.03511 34.71798 0.8055256
## Report 4-Report 1 -205.5585235 -260.31065 -150.80640 0.0000000
## Report 3-Report 2 -0.7433636 -52.68605 51.19933 0.9999821
## Report 4-Report 2 -187.6433234 -240.99857 -134.28808 0.0000000
## Report 4-Report 3 -186.8999598 -240.37547 -133.42445 0.0000000## [1] "Level 1 Semester 2"## Df Sum Sq Mean Sq F value Pr(>F)
## df[, 9] 2 13075208 6537604 73.75 <2e-16 ***
## Residuals 2345 207879794 88648
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 307 observations deleted due to missingness## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = df[, 10] ~ df[, 9])
##
## $`df[, 9]`
## diff lwr upr p adj
## Report 2-Report 1 -12.21178 -46.7486 22.32503 0.6848886
## Report 3-Report 1 -167.92481 -203.4405 -132.40910 0.0000000
## Report 3-Report 2 -155.71302 -191.8087 -119.61734 0.0000000How long do students pause to look at feedback/their writing (between clicks)?
Minimum (in msec), and Max (in hours) as tables
## BIOL1040Sem1Report 1 BIOL1040Sem1Report 2 BIOL1040Sem1Report 3
## 82 35 18
## BIOL1040Sem1Report 4 BIOL1040Sem2Report 1 BIOL1040Sem2Report 2
## 40 7 16
## BIOL1040Sem2Report 3 BIOM2011Sem1Report 1 BIOM2011Sem1Report 2
## 11 129 476
## BIOM2011Sem2Report 1 BIOM2011Sem2Report 2
## 96 446## BIOL1040Sem1Report 1 BIOL1040Sem1Report 2 BIOL1040Sem1Report 3
## 23.6 23.8 45.8
## BIOL1040Sem1Report 4 BIOL1040Sem2Report 1 BIOL1040Sem2Report 2
## 21.7 36.3 31.3
## BIOL1040Sem2Report 3 BIOM2011Sem1Report 1 BIOM2011Sem1Report 2
## 19.3 22.5 17.4
## BIOM2011Sem2Report 1 BIOM2011Sem2Report 2
## 17.2 9.7pause duration as mean and SEM
is there a sig diff in the pauses between reports?
## means errors
## Report 1 1.999917 0.1521152
## Report 2 1.270313 0.1893537## [1] "Level 2 - semesters pooled"##
## Welch Two Sample t-test
##
## data: dfs[, dv] by dfs[, iv]
## t = 3.0039, df = 31423.57, p-value = 0.002668
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 15212.20 72340.25
## sample estimates:
## mean in group Report 1 mean in group Report 2
## 119995.03 76218.81## [1] "Level 1 - semesters pooled"## Df Sum Sq Mean Sq F value Pr(>F)
## dfs[, iv] 3 1.347e+14 4.488e+13 11.48 1.61e-07 ***
## Residuals 314287 1.229e+18 3.911e+12
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = dfs[, dv] ~ dfs[, iv])
##
## $`dfs[, iv]`
## diff lwr upr p adj
## Report 2-Report 1 20610.105 -761.7298 41981.939 0.0634555
## Report 3-Report 1 -5627.407 -29855.8705 18601.056 0.9331029
## Report 4-Report 1 -69955.224 -110174.2824 -29736.165 0.0000466
## Report 3-Report 2 -26237.512 -51379.3957 -1095.627 0.0368834
## Report 4-Report 2 -90565.328 -131341.1615 -49789.495 0.0000001
## Report 4-Report 3 -64327.817 -106670.8411 -21984.792 0.0005506## [1] "Level 1 Semester 1"## Df Sum Sq Mean Sq F value Pr(>F)
## dfs[, iv] 3 1.514e+14 5.047e+13 11.67 1.22e-07 ***
## Residuals 166354 7.196e+17 4.326e+12
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = dfs[, dv] ~ dfs[, iv])
##
## $`dfs[, iv]`
## diff lwr upr p adj
## Report 2-Report 1 -1850.813 -35629.93 31928.31 0.9990057
## Report 3-Report 1 1827.874 -32317.63 35973.38 0.9990724
## Report 4-Report 1 -96379.066 -142556.58 -50201.56 0.0000005
## Report 3-Report 2 3678.688 -30453.29 37810.67 0.9925910
## Report 4-Report 2 -94528.253 -140695.76 -48360.74 0.0000009
## Report 4-Report 3 -98206.940 -144643.19 -51770.69 0.0000003## [1] "Level 1 Semester 1"## Df Sum Sq Mean Sq F value Pr(>F)
## dfs[, iv] 2 1.592e+14 7.962e+13 23.12 9.12e-11 ***
## Residuals 147930 5.094e+17 3.443e+12
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = dfs[, dv] ~ dfs[, iv])
##
## $`dfs[, iv]`
## diff lwr upr p adj
## Report 2-Report 1 34371.05 9691.028 59051.07 0.0031523
## Report 3-Report 1 -70504.64 -105281.447 -35727.83 0.0000060
## Report 3-Report 2 -104875.69 -141164.763 -68586.62 0.0000000pause duration as histograms
NB log scale, so
## -4 -3 -2 -1 0 1 2 3
## "1 sec" "3 sec" "8 sec" "22 sec" "1 min" "3 min" "7 min" "20 min"
## 4 5 6 7 8
## "55 min" "2 hr" "7 hr" "18 hr" "50 hr"
setting up final notfinal categorisation