Merging some differences from the answers.

Author: jreades

Practical-09-Correlation.ipynb

@@ -1456,7 +1456,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "If you've finally managed to get it to all come together, your map should look something like this (broadly, that's actually the result of a _1:Many_ (_1:n_) join since each LSOA has in a given borough has the same value whereas our example now only contains boroughs, not LSOAs:\n",
+    "If you've managed to get it to all come together, your map should look something like the one below, though here you can see that it's actually the result of a _1:Many_ (_1:n_) join since each LSOA in a given borough has been given the same value whereas our example now only contains boroughs, not LSOAs:\n",
     "\n",
     "![PySAL Logo](https://kingsgeocomputation.files.wordpress.com/2017/11/cordf_r_map.png)"
    ]
@@ -1485,18 +1485,18 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "### Exercises\n",
+    "## Additional Exercises\n",
     "\n",
-    "Finally, here are some exercises to help you to reinforce, and extent upon, what you have learned above and throughout the module. "
+    "Finally, here are some exercises to help reinforce what you have learned above... and throughout the module. "
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "**Exercise 1**\n",
+    "#### Task 8.1\n",
     "\n",
-    "Build on the (looping) code for loopingcalculating correlations for all boroughs to do similar, but instead of calculating correlations, fit regressions (for specified variables in a DF) for ALL boroughs, summarising the results in a table. You may find this [SO question and answer](https://stackoverflow.com/questions/24088439/how-to-apply-ols-from-statsmodels-to-groupby) useful to guide you. "
+    "Build on the (looping) code for calculating correlations for all boroughs, but instead of calculating correlations you should fit regressions (for specified variables in a df) for ALL boroughs and summarise the results in a table. You may find this [SO question and answer](https://stackoverflow.com/questions/24088439/how-to-apply-ols-from-statsmodels-to-groupby) useful as a guide. "
    ]
   },
   {
@@ -1512,9 +1512,9 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "**Excercise 2**\n",
+    "#### Task 8.2\n",
     "\n",
-    "You may have noticed above that we repeated quite a lot of code, but with slight variations in object names, when fitting and analysing the regression models. In circumstances like that (when you are repeating code), it can he useful to write yourself a 'helper function' to speed up your analysis. \n",
+    "You may have noticed above that we repeated quite a lot of code, but with slight variations in variable and object names, when fitting and analysing the regression models. In circumstances like that (when you are repeating code), it can he useful to write yourself a 'helper function' to speed up your analysis. \n",
     "\n",
     "For this exercise, write a helper function to: \n",
     "- read a statsmodels.api OLS model and the data it uses\n",