Q1_1:The population mean number of flatworms is the same for each season. There is no effect of SEASON on the number of flatworms
Q1_2:There is no added variance due to SITES
Q1_2_1:Normality
Q1_2_2:Boxplots. Remember that it is that the replicates for the treatment that need to be normally distributed and thus make sure it is the correct residuals that are used to generate the boxplots.
Q1_2_3:Homogeneity of variance
Q1_2_4:Boxplots, plot of mean vs variance, residual plot. Consider which are the appropriate replicates (residuals) for each test.
Q1_2_5:Independent observations
Q1_2_6:Obeservations within the nesting factor still need to be as independent as posible - random, unbiased.
Q1_3:Y
Q1_4_1:SEASON, 1
Q1_4_2:SITE, 4
Q1_4_3:SITE, 4
Q1_4_4:Residual, 30
Q1_6a:1
Q1_6b:5.5709
Q1_6c:34.496
Q1_6d:0.004198
Q1_6e:4
Q1_6f:0.1615
Q1_6g:1.0635
Q1_6h:0.3917
Q1_6i:30
Q1_6j:0.1519
Q1_7:Statistical - Reject the null hypothesis that the population mean forth-root number of Dugesia was the same in each season. There was no added variability due to sites. Biological - There were significantly more Dugesia encounterd in winter than in summer. The number of Dugesia was not found to differ significantly across sites.
Q1_8:seasons
Q1_9a:In future, it might be more effective to ignore the impact of sites and use completely randomised designs
Q1_9b:If a nested design was to be used, then since SITES are the replicates for the effect of SEASON, then power is maximised by having more sites. Whilst having more stones may increase the precision of the measure of Dugusia within a site, it will not improve the power of the test of season.
Q2_1:That the population mean mass of starlings is the same for each roosting situation treatment - no effect of roosting situation on the population mean mass of starlings
Q2_1a:That the population mean mass of starlings is the same for each month - no effect of month on the population mean mass of starlings
Q2_1b:That the population mean mass of starlings is the same for each month and situation combination - no interaction between roosting situation and month on the population mean mass starlings
Q2_2a:N
Q2_3aa:N
Q2_3a:574.40
Q2_3b:3
Q2_3c:191.47
Q2_3d:10.8207
Q2_3e:<0.001
Q2_3f:1656.20
Q2_3g:1
Q2_3h:1656.20
Q2_3i:93.6000
Q2_3j:<0.001
Q2_3k:34.20
Q2_3l:3
Q2_3m:11.40
Q2_3n:0.6443
Q2_3o:0.5891
Q2_3p:1274.00
Q2_3q:72
Q2_3r:17.69
Q2_5a:NS
Q2_5b:NS
Q2_5c:*
Q2_5d:NS
Q2_5e:*
Q2_5f:NS
Q2_7:Starlings were found to be significantly heavier in January than in November. Irrespective of month, there was a significant difference in mean mass of birds roosting under different situations. Specifically, birds roosting in situation 1 and situation 2 were found to be significantly heavier than birds in situation 4.
Q3_1:N
Q3_2_1:N
Q3_2a:2.9293
Q3_2b:2
Q3_2c:1.4646
Q3_2d:5.7099
Q3_2e:0.004644
Q3_2f:2.4273
Q3_2g:1
Q3_2h:2.4273
Q3_2i:9.4629
Q3_2j:0.002786
Q3_2k:0.1914
Q3_2l:2
Q3_2m:0.0957
Q3_2n:0.3731
Q3_2o:0.689691
Q3_2p:22.8293
Q3_2q:89
Q3_2r:0.2565
Q3_4a:*
Q3_4b:NS
Q3_4c:NS
Q3_6:The rate of bud emergence was found to be significantly greater in healthy spruce plants than diseased spruce plants. Irrespective of disease status, soil pH was also found to have a significant affect on plant bud emergence rating. Plants grown in pH 3 soil were found to have significantly greater rates of bud emergence than plants grown in ph 5.5 soil.
Q3_7:The replicates for the anaylsis was the average bud emergence rating (average of 5 buds). An average based on five observations is going to be more precise than a single observation. Therefore, samples based on such averages should represent the populations better.
Q4_1_1:N
Q4_1a:87.454
Q4_1b:3
Q4_1c:29.151
Q4_1d:28.5740
Q4_1e:<0.001
Q4_1f:7.089
Q4_1g:1
Q4_1h:7.089
Q4_1i:6.9491
Q4_1j:0.01255
Q4_1k:11.354
Q4_1l:3
Q4_1m:3.785
Q4_1n:3.7098
Q4_1o:0.02068
Q4_1p:34.687
Q4_1q:34
Q4_1r:1.020
Q4_3a:N
Q4_3b:Mussel bed density was found to have a significant negative effect on mussel recruitment during summer, however, it was not found to alter recruitment in any other season (autumn, spring or winter).
Q5_1:The population mean number of lesions is the same for each treatment within leaf blocks. There is no effect of Treatment within leaf blocks
Q5_2:The population mean number of lesions is the same for each leaf block. There is no effect of leaf blocks
Q5_2_1:Normality
Q5_2_2:Boxplots
Q5_2_3:Homogeneity of variance
Q5_2_4:Boxplots, plot of means vs variance, residual plot
Q5_2_5:No interaction between blocking factor and the within block factor
Q5_2_6:Interaction plot
Q5_2_5:Independent observations
Q5_2_6:Observations within each block should be otherwise independent, unbiased. Randomize location of blocks and treatments within blocks.
Q5_3_6:The interaction plot suggests that there is some evidence of an interaction. Although the number of leasions appear to be greater in strongly innoculated leaves than the weakly inocculated leaves in most of the leaf pairs (blocks), this trend is either absent or reversed in two of the eight (1/4) of the leaf pairs. As a result, the test of block may not be reliable, and the power of the main test of treatment is reduced.
Q5_4a:7
Q5_4b:41.727
Q5_4e:1
Q5_4f:248.339
Q5_4g:17.172
Q5_4h:0.004328
Q5_4i:7
Q5_4j:14.462
Q5_5:Strongly innoculated tobacco leaves were found to have significantly higher mean numbers of lesions than weakly inoculated leaves. Leaf pairs (blocks) explained substantial amounts of the variation and therefore probably contributed to the sensitivity of the main test of treatment - thereby justifying the blocking design over a completely randomised design.