\documentclass[12pt]{article}
\usepackage{color}
\addtolength{\oddsidemargin}{-3cm}
\addtolength{\topmargin}{-2cm}
\setlength{\textwidth}{180mm}
\setlength{\textheight}{260mm}
\input{rgb}
\pagestyle{empty}
\begin{document}
\LARGE
\begin{itemize}
\item {\color{red} We fitted the flux density survey data
using a Bayesian polynomial regression analysis which assessed the PDF for
the required order of a polynomial fit.}
\end{itemize}

We chose the independent variable to be
\begin{equation}
{\color{blue} x = \log_{10} (\nu/\rm MHz) }
\end{equation}

and expanded the dependent variable 
{\color{blue}
\begin{equation}
y = \log_{10} S_{\nu} = \sum_{r=0}^{N} a_{r} x^{r}.
\end{equation}
}

\begin{itemize}
\item {\color{HotPink3} In a number of cases the peak in the {\sc pdf} led
us to prefer a first-order polynomial, and then the {\color{blue} spectral
index is $-a_{1}$}}.
\end{itemize}

\begin{itemize}
\item In many cases the {\sc pdf} led us to prefer a second-order fit,
implying significant curvature: \newline {\color{blue} spectral index at
{\color{purple3} rest frame} 151 MHz is $-a_{1} - 2a_{2}\log_{10}(151)$}.
\end{itemize}

\begin{itemize}
\item In no case was a higher order fit preferred.
\end{itemize}

\begin{itemize}
\item Luminosties were obtained for 
{\color{LimeGreen} (1) $\Omega_{\rm M} = 0$ and $\Omega_\Lambda = 0$}, \newline
{\color{blue}   (2) $\Omega_{\rm M} = 1$ and $\Omega_\Lambda = 0$} and 
{\color{LimeGreen} (3) $\Omega_{\rm M} = 0.1$ and $\Omega_\Lambda = 0.9$}.
\end{itemize}

\begin{itemize}
\item Tests were performed on just the certain FR\,IIs and also for
the combined set of FR\,IIs and DAs. No significant difference was
found.
\end{itemize}
\end{document}