Add citations for datasets

README.Rmd

@@ -150,7 +150,7 @@ pheatmap(
 
 ### Handling mixed variable types
 
-The `cor_polychoric()` function is currently limited in its flexibility as it only provides polychoric estimation for ordinal variables and does not support biserial or polyserial estimation for mixed ordinal and continuous variables. The function does, however, attempt to recognise potentially non-discrete variables, allowing for up to 10 levels, like in [World Values Survey](https://www.worldvaluessurvey.org/wvs.jsp) questionnaire items. In comparison, the `polychoric()` function from the `psych` package allows up to 8 levels by default.
+The `cor_polychoric()` function is currently limited in its flexibility as it only provides polychoric estimation for ordinal variables and does not support biserial or polyserial estimation for mixed ordinal and continuous variables. The function does, however, attempt to recognise potentially non-discrete variables, allowing for up to 10 levels, like in [World Values Survey](https://www.worldvaluessurvey.org/wvs.jsp) [@gedeshi2021] questionnaire items. In comparison, the `polychoric()` function from the `psych` package allows up to 8 levels by default.
 
 It's worth noting that variables with a high number of distinct values may cause estimation issues, so the `cor_polychoric()` function returns Spearman's $\rho$ instead (with a warning).
 
@@ -176,7 +176,7 @@ Due to its strong bivariate normality assumptions, `cor_polyserial()` is curren
 
 The `cor_polychoric()` function always uses pairwise complete observations. Therefore, the user need not worry about missing data. However, depending on the analysis design and the ratio of missing data, it may be essential to check for patterns of missingness and consider imputation.
 
-The General Social Survey Schwartz Values Module dataset is cleared of missing values (non-response or non-applicable). Here we introduce some NAs into random places across the dataset. The summary will show the dataset now actually contains missings.
+The General Social Survey Schwartz Values Module dataset [@smith2014] is cleared of missing values (non-response or non-applicable). Here we introduce some NAs into random places across the dataset. The summary will show the dataset now actually contains missings.
 
 ```{r example-missing-01, set.seed(111)}
 gss_miss <- gss12_values
@@ -239,4 +239,3 @@ Please report any issues you came up with on the [issues](https://github.com/Mar
 1.  Implement (optional) more robust distributional assumptions, e.g. a skew normal distribution ([@jin2016]).
 
 </details>
-

README.md

@@ -378,8 +378,9 @@ not support biserial or polyserial estimation for mixed ordinal and
 continuous variables. The function does, however, attempt to recognise
 potentially non-discrete variables, allowing for up to 10 levels, like
 in [World Values Survey](https://www.worldvaluessurvey.org/wvs.jsp)
-questionnaire items. In comparison, the `polychoric()` function from the
-`psych` package allows up to 8 levels by default.
+(Gedeshi et al. 2021) questionnaire items. In comparison, the
+`polychoric()` function from the `psych` package allows up to 8 levels
+by default.
 
 It’s worth noting that variables with a high number of distinct values
 may cause estimation issues, so the `cor_polychoric()` function returns
@@ -419,10 +420,11 @@ However, depending on the analysis design and the ratio of missing data,
 it may be essential to check for patterns of missingness and consider
 imputation.
 
-The General Social Survey Schwartz Values Module dataset is cleared of
-missing values (non-response or non-applicable). Here we introduce some
-NAs into random places across the dataset. The summary will show the
-dataset now actually contains missings.
+The General Social Survey Schwartz Values Module dataset (Smith,
+Marsden, and Hout 2014) is cleared of missing values (non-response or
+non-applicable). Here we introduce some NAs into random places across
+the dataset. The summary will show the dataset now actually contains
+missings.
 
 ``` r
 gss_miss <- gss12_values
@@ -540,6 +542,15 @@ Drasgow, Fritz. 2004. “Polychoric and Polyserial Correlations,” October.
 
 </div>
 
+<div id="ref-gedeshi2021" class="csl-entry">
+
+Gedeshi, Ilir, Merab Pachulia, David Rotman, Sylvia Kritzinger, Georg
+Poghosyan, Georgy Fotev, Jadranka Kolenović-apo, et al. 2021. “Joint
+EVS/WVS 2017-2021 Dataset (Joint EVS/WVS).” World Values Survey
+Association. <https://doi.org/10.14281/18241.11>.
+
+</div>
+
 <div id="ref-jin2016" class="csl-entry">
 
 Jin, Shaobo, and Fan Yang-Wallentin. 2016. “Asymptotic Robustness Study
@@ -571,4 +582,13 @@ Psychometric, and Personality Research.”
 
 </div>
 
+<div id="ref-smith2014" class="csl-entry">
+
+Smith, Tom W., Peter V. Marsden, and Michael Hout. 2014. “General Social
+Survey, 2012 Merged Data, Including a Cultural Module \[United
+States\].” ICPSR - Interuniversity Consortium for Political; Social
+Research. <https://doi.org/10.3886/ICPSR35478.V4>.
+
+</div>
+
 </div>

references.bib

@@ -100,3 +100,25 @@ @misc{eigenweb
     howpublished = {http://eigen.tuxfamily.org},
     year = {2010}
 }
+
+@misc{smith2014,
+	title = {General Social Survey, 2012 Merged Data, Including a Cultural Module [United States]},
+	author = {Smith, Tom W. and Marsden, Peter V. and Hout, Michael},
+	year = {2014},
+	date = {2014},
+	publisher = {ICPSR - Interuniversity Consortium for Political and Social Research},
+	doi = {10.3886/ICPSR35478.V4},
+	url = {https://www.icpsr.umich.edu/web/NADAC/studies/35478/versions/V4},
+	langid = {en}
+}
+
+@misc{gedeshi2021,
+	title = {Joint EVS/WVS 2017-2021 Dataset (Joint EVS/WVS)},
+	author = {Gedeshi, Ilir and Pachulia, Merab and Rotman, David and Kritzinger, Sylvia and Poghosyan, Georg and Fotev, Georgy and {Kolenovi{\'{c}}-{\DJ}apo}, Jadranka and Balobana, Stjepan and Baloban, Josip and {Rabu{\v{s}}ic}, Ladislav and Frederiksen, Morten and Saar, Erki and Ketola, Kimmo and Pachulia, Merab and {Bréchon}, Pierre and Wolf, Christof and Rosta, Gergely and Voas, David and Rovati, Giancarlo and {Jónsdóttir}, {Guðbjörg A.} and Ziliukaite, Ruta and Petkovska, Antoanela and Reeskens, Tim and Komar, Olivera and Jenssen, Anders T. and Voicu, Bogdan and Soboleva, Natalia and Marody, {Miros{\l}awa} and {Be{\v{s}}i{\'{c}}}, {Milo{\v{s}}} and {Strapcová}, Katarina and Uhan, Samo and Silvestre Cabrera, {María} and {Wallman-Lundåsen}, Susanne and {Ernst Stähli}, {Michèle} and Ramos, Alice and {Micó Ibáñez}, Joan and Carballo, Marita and McAllister, Ian and {Foa, Roberto Stefan (PI Bangladesh)} and Moreno Morales, Daniel E. and De Oliveira De Castro, Henrique Carlos and Lagos, Marta and Zhong, Yang and Casas, Andres and Yesilada, Birol and Paez, Cristina and Abdel Latif, Abdel Hamid and Jennings, Will and Welzel, Christian and Koniordos, Sokratis and {Díaz Argueta}, {Julio César} and Cheng, Edmund and Gravelle, Timothy and Stoker, Gerry and Dagher, Munqith and Yamazaki, Seiko and Braizat, Fares and Rakisheva, Botagoz and Bakaloff, Yuri and Haerpfer, Christian and Wing-Yat Yu, Eilo and Lee, Grace and Moreno, Alejandro and Souvanlasy, Chansada and Perry, Paul and Denton, Carlos and Puranen, Bi and Gilani, Bilal and Romero, Catalina and Guerrero, Linda and {Hernández Acosta}, Javier J. and Voicu, Bogdan and Zavadskaya, Margarita and Veskovic, Nino and Auh, Soo Young and Tsai, Ming-Chang and Olimov, Muzaffar and Bureekul, Thawilwadee and Ben Hafaiedh, Abdelwahab and Esmer, Yilmaz and Inglehart, Ronald and Depouilly, Xavier and Norris, Pippa and Balakireva, Olga},
+	year = {2021},
+	date = {2021},
+	publisher = {World Values Survey Association},
+	doi = {10.14281/18241.11},
+	url = {http://www.worldvaluessurvey.org/WVSEVSjoint2017.jsp},
+	langid = {en}
+}