This is a preregistration of an extension project for the following study:
The Pork Barrel Politics of the Towns Fund
Author: Chris Hanretty
Reference:
Hanretty, C. (2021) ‘The Pork Barrel Politics of the Towns Fund’, The Political quarterly (London. 1930), 92(1), pp. 7–13. doi: 10.1111/1467-923X.12970.
This replication project is focused on the following argument made in the study:
Claim: A town located in a seat with a Conservative majority of 0 to 5% voter margin will be 6.799 more likely to receive funding in the towns fund scheme.
Display item: Table 1, Model 4, Line 4
Describe scientific rationale for the study extension.
The extension of this replication project will add another level to the original Hanretty model. In the paper, the “Region” variable(which region a town is located in) was used as a control variable in the model. In this extension, it will be used as a level 2 variable instead. This will allow for the analysis of the original model to be varied by region and it will be possible to see how this relationship varies geographically. The same conservative majority categorical variable will be used as the first model. This will allow the examination of how the variable differs across different regions. The two models will then be compared to see if the new model with random slopes fits the data better then the original model created by Hanretty.
Hypothesis:
Model 4 from the Pork Barrel Politics of The Towns Fund paper will not fit the data as well as a Level 2 Random Slopes Model in a Bayesian Information Criterion Test. This is because funding from The Towns Fund will vary across regions of England.
I will test the hypothesis using the a different regression model to the one in the Pork Barrel Politics study. This model will differ from the original as the model will be now be a two level logit model with towns(level 1 variable) being nested in regions(level 2 variable). The model will still use the ConMaj.categorical variable with the factors used in the weighted formula used as controls.
Dependent variable(s):
*Outcome: Was the dependent variable in the original paper. Outcome is a binary variable showing which towns received funding from the government with the towns fund. The data was created from another variable in the set called funding. This was the same variable but was not on a numerical scale. A score of 0 means that that the town did not get funding. A score of 1 means that the town did get funding.
Independent variables (IVs):
The standard p<0.05 criteria will be used for determining if the statistical tests suggest that the results are significantly different from those expected from the null hypothesis.
No checks will be performed to determine eligibility for inclusion in this data set. This is unnecessary as the data was manipulated already by its creator to have all the necessary data for each town. Outliers will be included in the analysis.
State your process for dealing with missing data or state not applicable.
This section is non-applicable to the study as there is no missing data for the variables used in the set.
Since the model is a two level model with towns nested in regions. A visualisation of how many towns are nestled in regions would be appropriate. The bars will also be feature the colours of how many towns were in Tory seats compared to how many were in seats held by another political party.
A line graph showing the relationship between the ConMaj.allm variable, the non-categorical version of the conservative majority variable, and the outcome variable will be used in to show how the relationship between having a conservative majority and receiving funding from the Towns Fund changes per region.
This preregistration form was completed in the following R environment:
## R version 4.2.1 (2022-06-23 ucrt)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows 10 x64 (build 19044)
##
## Matrix products: default
##
## locale:
## [1] LC_COLLATE=English_United Kingdom.utf8
## [2] LC_CTYPE=English_United Kingdom.utf8
## [3] LC_MONETARY=English_United Kingdom.utf8
## [4] LC_NUMERIC=C
## [5] LC_TIME=English_United Kingdom.utf8
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] knitr_1.42 rmarkdown_2.21
##
## loaded via a namespace (and not attached):
## [1] digest_0.6.31 R6_2.5.1 jsonlite_1.8.4 evaluate_0.20
## [5] cachem_1.0.8 rlang_1.1.1 cli_3.6.1 rstudioapi_0.14
## [9] jquerylib_0.1.4 bslib_0.4.2 tools_4.2.1 xfun_0.39
## [13] yaml_2.3.7 fastmap_1.1.1 compiler_4.2.1 htmltools_0.5.5
## [17] sass_0.4.6Bartlett J. (2021). OSF preregistration template.Rmd. GitHub (accessed 03/05/2023)
Bowman, S. D., DeHaven, A. C., Errington, T. M., Hardwicke, T. E., Mellor, D. T., Nosek, B. A., & Soderberg, C. K. (2020). OSF Prereg Template. https://doi.org/10.31222/osf.io/epgjd. OSF (accessed 03/05/2023)