我有 2 个附录，B 和 C。我有几张图片和表格，希望将它们放在每个附录下。但是，尽管我尽了最大努力，但插入了所有相关信息后，所有图片和表格都杂乱无章，顺序也不正确。表 B1 和 B2 正确显示在附录 B 下，但是图 B3 到 B10 虽然按正确顺序显示，并且格式正确，却显示在 C1 到 C3 之后。表 C4 显示在 B3 到 B10 之后。如果有人能检查我的代码并找出我哪里出错了，我将不胜感激。如果需要更多信息，我很乐意提供。非常感谢
\documentclass[12pt]{article}
\usepackage[utf8]{inputenc}
\usepackage{fullpage}
\usepackage{british}
\usepackage{graphicx}
\usepackage{natbib}
\usepackage{appendix}
\usepackage{subcaption}
\usepackage{afterpage}
\usepackage{adjustbox}
\usepackage[font=small,labelfont=bf]{caption}
\usepackage{amsmath}
\usepackage{float}
\usepackage{caption,subcaption}
\begin{document}
\tableofcontents
\section{Appendix}
\begin{appendices}
\begin{table}[!hbp]
\caption* {Table B1: Variable Definitions} \label{tab:title}
\centering
\begin{adjustbox}{width=1\textwidth}
\begin{tabular}{l|l}
\textbf{Variable} & \textbf{Definition} \\ \hline
PRICE18 & IPL 2018 auction price. \\
LNPRICE18 & Natural logarithm of IPL 2018 auction price. \\
RESPRICE18 & Reservation IPL 2018 auction price of a player. \\
AGE & The age of a player, measured in years as of 26th January 2018. \\
ODIEXP & The number of One Day International (ODI) matches played. \\
ODIBATSR & ODI Batting Strike Rate. Calculated as the no. of runs per ball multiplied by 100. \\
ODIBOWLAV & ODI Bowling Average. Calculated as the no. of runs conceded by a bowler per wicket. \\
ODIBATRANK & ICC ODI Batting Rank. \\
T20IBATAV & T20I Batting Average. Calculated as the total no. of runs per total no. of overs faced. \\
T20IBOWLSR & ODI Bowling Strike Rate. Calculated as the no. balls bowled per wicket. \\
T20IBOWLRANK & ICC T20I Bowling Rank. \\
IPLEXP & The number of Indian Premier League (IPL) matches played \\
IPLBOWLAV & IPL Bowling Average. Calculated as the no. of runs conceded by a bowler per wicket. \\
IPLECON & IPL Economy Rate. Calculated as the no. of runs conceded per over bowled. \\
IPLFIELD & Total no. of IPL Fielding Dismissals \\ \hline
IND & Dummy variable: taking the value 1 if the player is Indian, 0 otherwise \\
AUS & Dummy variable: taking the value 1 if the player is Australian, 0 otherwise \\
SA & Dummy variable: taking the value 1 if the player is South African, 0 otherwise \\
ALL & Dummy variable: taking the value 1 if the player is an All-Rounder, 0 otherwise \\
FAMOUS & \begin{tabular}[c]{@{}l@{}} Dummy variable: taking the value 1 if the player has more than 1 million \\Instagram followers, 0 otherwise \end{tabular} \\
FORLEAGUE & \begin{tabular}[c]{@{}l@{}} Dummy variable: taking the value 1 if the player has played in a T20 tournament \\other than the IPL , 0 otherwise \end{tabular} \\
SOLD & Dummy variable: taking the value 1 if the player was sold at auction, 0 otherwise
\end{tabular}
\end{adjustbox}
\end{table}
\begin{table}[!hbp]
\caption* {Table B2: Summary Statistcs} \label{tab:title}
\centering
\begin{adjustbox}{width=0.8\textwidth}
\begin{tabular}{l|lllll}
\textbf{Variable} & \textbf{Observations} & \textbf{Mean} & \textbf{Std. Dev.} & \textbf{Min} & \textbf{Max} \\ \hline
LNPRICE18 & 186 & 5.0752 & 1.3618 & 3 & 7.44 \\
PRICE18 & 186 & 331.5 & 356.4 & 20 & 1700 \\
AGE & 312 & 26.69 & 4.726 & 16 & 39 \\
FAMOUS & 312 & 0.05448 & .2273 & 0 & 1 \\
IND & 312 & 0.6346 & 0.4823 & 0 & 1 \\
AUS & 312 & 0.1186 & 0.3238 & 0 & 1 \\
BAT & 312 & 0.2051 & 0.4044 & 0 & 1 \\
ALL & 312 & 0.3526 & 0.4785 & 0 & 1 \\
ODIEXP & 312 & 30.19 & 55.87 & 0 & 312 \\
ODIBATSR & 312 & 39.96 & 47.10 & 0 & 250 \\
ODIBOWLAV & 312 & 15.20 & 25.37 & 0 & 172 \\
ODIBATRANK & 312 & 58.33 & 15.59 & 1 & 65 \\
T20IBATAV & 312 & 7.890 & 11.87 & 0 & 52.86 \\
T20IBOWLSR & 312 & 7.269 & 11.42 & 0 & 68 \\
T20IBOWLRANK & 312 & 53.52 & 13.58 & 1 & 59 \\
IPLEXP & 312 & 27.42 & 41.29 & 0 & 191 \\
IPLBOWLAV & 312 & 12.64 & 18.29 & 0 & 90 \\
IPLECON & 312 & 3.185 & 4.064 & 0 & 14.2 \\
IPLFIELD & 312 & 9.087 & 16.64 & 0 & 90 \\
RESPRICE & 312 & 139.2 & 241.4 & 2 & 1700 \\
FORLEAGUE & 312 & 0.3397 & 0.4744 & 0 & 1 \\
SOLD & 312 & 0.5961 & 0.4915 & 0 & 1
\end{tabular}
\end{adjustbox}
\end{table}
\newpage
\begin{figure}[ht]
\centering
\setkeys{Gin}{width=1\textwidth}
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B3}
\includegraphics{example image 1}
\end{subfigure}%
\hfill
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B4}
\includegraphics{example image 2}
\end{subfigure}
\bigskip
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B5}
\includegraphics{example image 3}
\end{subfigure}%
\hfill
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B6}
\includegraphics{example image 4}
\end{subfigure}
\bigskip
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B7}
\includegraphics{example image 5}
\end{subfigure}%
\hfill
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B8}
\includegraphics{example image 6}
\end{subfigure}
\end{figure}
\bigskip
\begin{figure}[htb!]
\centering
\begin{subfigure}[htb!]{1\textwidth}
\caption*{Figure B9}
\includegraphics[width=1\textwidth]{example image 7}
\caption*{*figure (right) shows the positive correlation for the upper tail of player valuations.}
\end{subfigure}
\end{figure}
\begin{figure}[htb!]
\centering
\begin{subfigure}[htb!]{0.5\textwidth}
\caption*{Figure B10}
\includegraphics[width=1\textwidth]{example image 8}
\LARGE
\caption*{*Outliers are marked in red with Mandeep Singh only having played a total of 3 T20I matches and Manoj Tiwary playing his last T20I in 2012.}
\end{subfigure}
\end{figure}
\vfill
\section{Appendix}
\newpage
\begin{table}[htb!]
\centering
\caption*{\label{tab:table-name}\textbf{Table C1: Heteroskedasticity Tests}}
\begin{tabular}{l|l}
\multicolumn{1}{c|}{\textbf{OLS ICC Specification}} & \multicolumn{1}{c}{\textbf{OLS Raw Statistic Specification}} \\ \hline
Breusch-Pagan/ Cook-Weisberg test* & Breusch-Pagan/ Cook-Weisberg test* \\ \hline
H0 : Constant Variance (Homoskedastic) & H0 : Constant Variance (Homoskedastic) \\ \hline
$\chi$ ${}^{2\ }$ Test statistic = 9.42 & $\chi$ ${}^{2\ }$ Test statistic = 13.61 \\
Probability = 0.0021 & Probability = 0.0002
\end{tabular}
\caption*{\label{tab:table-name}* The Breush-Pagan/Cook-Weisberg Test were carried out on the normal standard specifications. The null hypothesis of constant variance or homoskedastic can be rejected for both specifications.}
\end{table}
\begin{table}[htb!]
\centering
\caption*{\label{tab:table-name}\textbf{Table C2: Misspecification Tests}}
\begin{tabular}{l|l}
\multicolumn{1}{c|}{\textbf{OLS ICC Specification}} & \multicolumn{1}{c}{\textbf{OLS Raw Statistic Specification}} \\ \hline
Ramsey RESET test* & Ramsey RESET test* \\ \hline
H0 : Model has no omitted variables & H0 : Model has no omitted variables \\ \hline
F (3, 169) = 12.64 & F (3, 169) = 8.82 \\
Probability = 0.0000 & Probability = 0.0000
\end{tabular}
\caption*{\label{tab:table-name}* The Ramsay Regression Equation Specification Error Test (RESET) tests whether non-linear combinations of the fitted values help explain the dependant variable, ‘lnprice18’. The null hypothesis of no omitted variables can be rejected for both specifications. }
\end{table}
\begin{table}[htb!]
\centering
\caption*{\label{tab:table-name}\textbf{Table C3: Tests for Endogeneity of the reserve price}}
\begin{tabular}{lll}
\multicolumn{3}{l}{\textbf{Regression of generalised residuals on the reserve price*}} \\ \hline
\multicolumn{1}{l|}{} & GRESICCNR & GRESRAWNR \\ \hline
\multicolumn{1}{l|}{RESPRICE18} & 0.0037 & 0.0036 \\
\multicolumn{1}{l|}{CONST} & 4.37 & 4.42 \\
\multicolumn{1}{l|}{NO. OF OBSERVATIONS} & 312 & 312 \\ \hline
\end{tabular}
\caption*{\label{tab:table-name}* The dependant variables are the generalised residuals from the OLS ICC & Raw Statistic Specifications, respectively. These residuals are predicted based on each specification excluding the reservation price as an explanatory variable. The statistical significance, at the 1\% level, imply the reservation price of auctioned players is highly correlated with the residuals from the estimated models. As a result, a null hypothesis of no correlation between the estimated error and the reserve price can be rejected for both specifications. }
\end{table}
\begin{table}[htb!]
\centering
\begin{tabular}{llllll}
\multicolumn{3}{c}{\textbf{OLS ICC Specification}} & \multicolumn{3}{c}{\textbf{OLS Raw Statistic Specification}} \\ \hline
\multicolumn{3}{l|}{\textbf{Correlation Matrix*}} & \multicolumn{3}{l}{\textbf{Correlation Matrix*}} \\ \hline
\multicolumn{1}{l|}{} & GRESICCNR & \multicolumn{1}{l|}{GRESICCR} & \multicolumn{1}{l|}{} & GRESRAWNR & GRESRAWR \\ \hline
\multicolumn{1}{l|}{GRESICCNR} & 1 & \multicolumn{1}{l|}{} & \multicolumn{1}{l|}{GRESRAWNR} & 1 & \\
\multicolumn{1}{l|}{GRESICCR} & \textbf{0.8905} & \multicolumn{1}{l|}{1} & \multicolumn{1}{l|}{GRESRAWR} & \textbf{0.9235} & 1 \\ \hline
\end{tabular}
\caption*{\label{tab:table-name}* The matrices show the correlation between the generalised residuals of the OLS specifications excluding the reserve price as an explanatory variable, ‘gresiccnr’ (‘gresrawnr’) and those when regressing the reserve price on all variables, except the reserve price itself. These results strengthen the case for base price being an endogenous variable.}
\end{table}
\begin{table}[htb!]
\centering
\caption*{\label{tab:table-name}\textbf{Table C4: Correlation between residuals of Retained and Non-Retained Players}}
\begin{adjustbox}{width=1\textwidth}
\begin{tabular}{llllll}
\multicolumn{3}{c}{\textbf{OLS ICC Specification}} & \multicolumn{3}{c}{\textbf{OLS Raw Statistic Specification}} \\ \hline
\multicolumn{3}{l}{\textbf{Correlation Matrix*}} & \multicolumn{3}{l}{\textbf{Correlation Matrix*}} \\ \hline
\multicolumn{1}{l|}{} & GRESICCNRT & \multicolumn{1}{l|}{GRESICCRT} & \multicolumn{1}{l|}{} & GRESRAWNRT & GRESRAWRT \\ \hline
\multicolumn{1}{l|}{GRESICCNRT} & 1 & \multicolumn{1}{l|}{} & \multicolumn{1}{l|}{GRESRAWNRT} & 1 & \\
\multicolumn{1}{l|}{GRESICCRT} & \textbf{0.9701} & \multicolumn{1}{l|}{1} & \multicolumn{1}{l|}{GRESRAWRT} & \textbf{0.9917} & 1
\\ \hline
\end{tabular}
\end{adjustbox}
\caption*{\label{tab:table-name}* The matrices show the correlation between the generalised residuals of the OLS specifications, ‘gresiccrt’ (‘gresrawrt’) and the residuals excluding the 17 retained players ‘gresiccnrt’ (‘gresrawnrt’). These residuals are found to be highly correlated suggesting there is some unmodelled factor affecting both retained and non-retained players. As a result, the valuations, for retained players, cannot be excluded from the estimated models without having an impact on the estimated effects of all other variables. }
\end{table}
\end{appendices}
\end{document}
答案1
下面的 MWE 旨在作为您编写附录的起点。由于我们没有您的图片，因此可能会发生某些图不适合页面的情况。在这种情况下，您可能应该重新组织图并更改其中的子图数量。您的图没有标题也是不寻常的（或者可能是我误解了，您喜欢用例如来表示）\caption*{Figure B10}。
\documentclass[12pt]{article}
\usepackage{fullpage} % better is to use `geometry`
\usepackage[demo,
export]{adjustbox}
\usepackage[font=small,
labelfont=bf]{caption}
\usepackage{subcaption}
\captionsetup[subfigure]{skip=0.25ex}
\usepackage{ragged2e} % new
\usepackage[referable,flushleft]{threeparttablex} % new
\usepackage{tabularx}
\newcolumntype{L}{>{\RaggedRight}X}
\usepackage{siunitx} % new
\usepackage{chngcntr} % new
\usepackage{etoolbox} % new
\AtBeginEnvironment{tablenotes}% new
{\footnotesize\linespread{0.84}\selectfont\RaggedRight}
\begin{document}
\counterwithin{figure}{section}
\counterwithin{table}{section}
\renewcommand{\thesection}{\Alph{section}}
\renewcommand{\thefigure}{\thesection\arabic{figure}}
\renewcommand{\thetable}{\thesection\arabic{table}}
\section{Appendix}
\begin{table}[!hbp]
\caption{Variable Definitions}
\label{tab:title}
\centering
\begin{tabularx}{\linewidth}{l|Xl}
\textbf{Variable} & \textbf{Definition} \\ \hline
PRICE18 & IPL 2018 auction price. \\
LNPRICE18 & Natural logarithm of IPL 2018 auction price. \\
RESPRICE18 & Reservation IPL 2018 auction price of a player. \\
AGE & The age of a player, measured in years as of 26th January 2018. \\
ODIEXP & The number of One Day International (ODI) matches played. \\
ODIBATSR & ODI Batting Strike Rate. Calculated as the no. of runs per ball multiplied by 100. \\
ODIBOWLAV & ODI Bowling Average. Calculated as the no. of runs conceded by a bowler per wicket. \\
ODIBATRANK & ICC ODI Batting Rank. \\
T20IBATAV & T20I Batting Average. Calculated as the total no. of runs per total no. of overs faced. \\
T20IBOWLSR & ODI Bowling Strike Rate. Calculated as the no. balls bowled per wicket. \\
T20IBOWLRANK & ICC T20I Bowling Rank. \\
IPLEXP & The number of Indian Premier League (IPL) matches played \\
IPLBOWLAV & IPL Bowling Average. Calculated as the no. of runs conceded by a bowler per wicket. \\
IPLECON & IPL Economy Rate. Calculated as the no. of runs conceded per over bowled. \\
IPLFIELD & Total no. of IPL Fielding Dismissals \\ \hline
IND & Dummy variable: taking the value 1 if the player is Indian, 0 otherwise \\
AUS & Dummy variable: taking the value 1 if the player is Australian, 0 otherwise \\
SA & Dummy variable: taking the value 1 if the player is South African, 0 otherwise \\
ALL & Dummy variable: taking the value 1 if the player is an All-Rounder, 0 otherwise \\
FAMOUS & Dummy variable: taking the value 1
if the player has more than 1 million
Instagram followers, 0 otherwise \\
FORLEAGUE & Dummy variable: taking the value 1 if the player has played in a T20 tournament other than the IPL , 0 otherwise \\
SOLD & Dummy variable: taking the value 1 if the player was sold at auction, 0 otherwise
\end{tabularx}
\end{table}
\begin{table}[!hbp]
\caption{Summary Statistics} \label{tab:title}
\centering
\begin{tabularx}{\linewidth}{@{}X|S[table-format=3]
S[table-format=3.4]
S[table-format=1.4]
S[table-format=2]
S[table-format=4.2]
@{}}
\textbf{Variable} & \textbf{Observ.} & \textbf{Mean} & \textbf{Std. Dev.} & \textbf{Min} & \textbf{Max} \\ \hline
LNPRICE18 & 186 & 5.0752 & 1.3618 & 3 & 7.44 \\
PRICE18 & 186 & 331.5 & 356.4 & 20 & 1700 \\
AGE & 312 & 26.69 & 4.726 & 16 & 39 \\
FAMOUS & 312 & 0.05448 & .2273 & 0 & 1 \\
IND & 312 & 0.6346 & 0.4823 & 0 & 1 \\
AUS & 312 & 0.1186 & 0.3238 & 0 & 1 \\
BAT & 312 & 0.2051 & 0.4044 & 0 & 1 \\
ALL & 312 & 0.3526 & 0.4785 & 0 & 1 \\
ODIEXP & 312 & 30.19 & 55.87 & 0 & 312 \\
ODIBATSR & 312 & 39.96 & 47.10 & 0 & 250 \\
ODIBOWLAV & 312 & 15.20 & 25.37 & 0 & 172 \\
ODIBATRANK & 312 & 58.33 & 15.59 & 1 & 65 \\
T20IBATAV & 312 & 7.890 & 11.87 & 0 & 52.86 \\
T20IBOWLSR & 312 & 7.269 & 11.42 & 0 & 68 \\
T20IBOWLRANK & 312 & 53.52 & 13.58 & 1 & 59 \\
IPLEXP & 312 & 27.42 & 41.29 & 0 & 191 \\
IPLBOWLAV & 312 & 12.64 & 18.29 & 0 & 90 \\
IPLECON & 312 & 3.185 & 4.064 & 0 & 14.2 \\
IPLFIELD & 312 & 9.087 & 16.64 & 0 & 90 \\
RESPRICE & 312 & 139.2 & 241.4 & 2 & 1700 \\
FORLEAGUE & 312 & 0.3397 & 0.4744 & 0 & 1 \\
SOLD & 312 & 0.5961 & 0.4915 & 0 & 1
\end{tabularx}
\end{table}
\begin{figure}[ht]
\setkeys{Gin}{width=\linewidth}
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B3}
\includegraphics{IPL Nationality Breakdown (Full Data Set).png}
\end{subfigure}%
\hfill
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B4}
\includegraphics{IPL Nationality Breakdown (SOLD).png}
\end{subfigure}
\medskip
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B5}
\includegraphics{graph1.png}
\end{subfigure}%
\hfill
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B6}
\includegraphics{graph2.png}
\end{subfigure}
\medskip
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B7}
\includegraphics{example image 5}
\end{subfigure}%
\hfill
\begin{subfigure}[t]{0.49\textwidth}
\caption*{Figure B8}
\includegraphics{IPL Bowling economy auction price.png}
\end{subfigure}
\medskip
\begin{subfigure}[t]{0.65\linewidth}
\caption*{Figure B9}
\includegraphics{insta.jpg}
\begin{tablenotes}
\note{figure (right) shows the positive correlation for the upper tail of player valuations.}
\end{tablenotes}
\end{subfigure}
\hfill
\begin{subfigure}[t]{0.34\textwidth}
\caption*{Figure B10}
\includegraphics{relationship t20i batav auction price.png}
\begin{tablenotes}
\note{*Outliers are marked in red with Mandeep Singh only having played a total of 3 T20I matches and Manoj Tiwary playing his last T20I in 2012.}
\end{tablenotes}
\end{subfigure}
\end{figure}
\clearpage
\section{Appendix}
\begin{table}[!ht]
\centering
\caption{Heteroskedasticity Tests}
\label{tab:B1}
\begin{tabularx}{\linewidth}{@{} L|L @{}}
\textbf{OLS ICC Specification} & \textbf{OLS Raw Statistic Specification} \\ \hline
Breusch-Pagan/ Cook-Weisberg test* & Breusch-Pagan/ Cook-Weisberg test* \\ \hline
H0 : Constant Variance (Homoskedastic) & H0 : Constant Variance (Homoskedastic) \\ \hline
$\chi$ ${}^{2\ }$ Test statistic = 9.42 & $\chi$ ${}^{2\ }$ Test statistic = 13.61 \\
Probability = 0.0021 & Probability = 0.0002
\end{tabularx}
\begin{tablenotes}
\note{The Breush-Pagan/Cook-Weisberg Test were carried out on the normal standard specifications. The null hypothesis of constant variance or homoskedastic can be rejected for both specifications.}
\end{tablenotes}
\end{table}
\begin{table}[!ht]
\begin{center}
\caption{Misspecification Tests}
\label{tab:B2}
\begin{tabular}{l|l}
\textbf{OLS ICC Specification}
& \textbf{OLS Raw Statistic Specification} \\ \hline
Ramsey RESET test* & Ramsey RESET test* \\ \hline
H0 : Model has no omitted variables
& H0 : Model has no omitted variables \\ \hline
F (3, 169) = 12.64 & F (3, 169) = 8.82 \\
Probability = 0.00 & Probability = 0.0000
\end{tabular}\vspace{-0.75\baselineskip}
\end{center}
\begin{tablenotes}
\note{The Ramsay Regression Equation Specification Error Test (RESET) tests whether non-linear combinations of the fitted values help explain the dependant variable, ‘lnprice18’. The null hypothesis of no omitted variables can be rejected for both specifications.}
\end{tablenotes}
\end{table}
\begin{table}[!ht]
\caption{Tests for Endogeneity of the reserve price}
\label{tab:B3}
\begin{center}
\begin{tabular}{l|ll}
\multicolumn{3}{l}{\textbf{Regression of generalised residuals on the reserve price*}} \\ \hline
& GRESICCNR & GRESRAWNR \\ \hline
RESPRICE18
& 0.0037 & 0.0036 \\
CONST
& 4.37 & 4.42 \\
NO. OF OBSERVATIONS
& 312 & 312 \\ \hline
\end{tabular}\vspace{-0.75\baselineskip}
\end{center}
\begin{tablenotes}
\note{The dependant variables are the generalised residuals from the OLS ICC \& Raw Statistic Specifications, respectively. These residuals are predicted based on each specification excluding the reservation price as an explanatory variable. The statistical significance, at the 1\% level, imply the reservation price of auctioned players is highly correlated with the residuals from the estimated models. As a result, a null hypothesis of no correlation between the estimated error and the reserve price can be rejected for both specifications.}
\end{tablenotes}
\end{table}
\clearpage
\section{Appendix}
\begin{table}[!ht]
\centering
\caption{Tests for Endogeneity of the reserve price}
\label{tab:C1}
\begin{tabularx}{\linewidth}{@{} Xll|Xll @{}}
\multicolumn{3}{c}{\textbf{OLS ICC Specification}} & \multicolumn{3}{c}{\textbf{OLS Raw Statistic Specification}} \\ \hline
%
\multicolumn{3}{c|}{\textbf{Correlation Matrix}} & \multicolumn{3}{c}{\textbf{Correlation Matrix}} \\ \hline
& (1) & (2) & & (3) & (4) \\ \hline
GRESICCNR (1) & 1 & & GRESRAWNR (3) & 1 & \\
GRESICCR (2) & \textbf{0.8905}
& & GRESRAWR (4) & \textbf{0.9235}
& 1 \\ \hline
\end{tabularx}
\begin{tablenotes}
\note{The matrices show the correlation between the generalised residuals of the OLS specifications excluding the reserve price as an explanatory variable, ‘gresiccnr’ (‘gresrawnr’) and those when regressing the reserve price on all variables, except the reserve price itself. These results strengthen the case for base price being an endogenous variable.}
\end{tablenotes}
\end{table}
\begin{table}[!ht]
\centering
\caption{Correlation between residuals of Retained and Non-Retained Players}
\label{tab:C2}
\begin{tabularx}{\linewidth}{@{} Xll|Xll @{}}
\multicolumn{3}{c}{\textbf{OLS ICC Specification}} & \multicolumn{3}{c}{\textbf{OLS Raw Statistic Specification}} \\ \hline
%
\multicolumn{3}{c|}{\textbf{Correlation Matrix}} & \multicolumn{3}{c}{\textbf{Correlation Matrix}} \\ \hline
& (1) & (2) & & (3) & (4) \\ \hline
GRESICCNRT (1) & 1 & & GRESRAWNR (3) & 1 & \\
GRESICCRT (2) & \textbf{0.9701}
& & GRESRAWR (4) & \textbf{0.9917}
& 1 \\ \hline
\end{tabularx}
\begin{tablenotes}
\note{The matrices show the correlation between the generalised residuals of the OLS specifications, ‘gresiccrt’ (‘gresrawrt’) and the residuals excluding the 17 retained players ‘gresiccnrt’ (‘gresrawnrt’). These residuals are found to be highly correlated suggesting there is some unmodelled factor affecting both retained and non-retained players. As a result, the valuations, for retained players, cannot be excluded from the estimated models without having an impact on the estimated effects of all other variables.}
\end{tablenotes}
\end{table}
\end{document}
MWE 生成了长达五页的文档。每个附录都从新的一页开始。