Lesson video

Hi, my name is Chloe, and I'm a geography field studies tutor.

Today's lesson in the fieldwork unit is all about analysing geographical data and then trying to create really great conclusions.

Let's get started.

By the end of today's lesson, you will be able to analyse geographical data and use your findings to create different types of conclusions.

There are some key words in this lesson, so let's just review those now.

First of all, a correlation.

This is where there's a change in one variable, and it appears to be paired against a change in another variable.

Empirical evidence is evidence that's collected through direct observation and experience rather than through logical thinking.

And finally, confirmation bias.

This is where we conclude what we believe to be true by basing our ideas on some, but not all, of the evidence that we have collected.

So the lesson has three parts.

Let's just have a quick look at these.

How do geographers analyse quantitative data? That's where we're gonna start.

Then we're going to look at how geographers analyse qualitative data.

And then finally, we're going to think about, how does data analysis lead to good conclusions? So let's start with that first one.

Many people wrongly believe that you have to be good at maths to successfully analyse quantitative data.

In fact, successful analysis in geography means that you understand what the numerical values mean.

You don't necessarily have to be able to be good at processing and actually manipulating the data.

The first stage in analysing quantitative data is to effectively describe it, and something that geographers use a lot is the GRaDE acronym, and this helps them to structure their description.

So going through the GRaDE acronym, it stands for, G is general trend, the Ra is range of data, the D is data quotes, you actually have a quotation, and E is where you're highlighting exceptions.

We're gonna be going going through these four different parts of the description by looking at the graph that you can see on the screen now.

So let's start with the general trend.

Before we do, let's just have a quick look at the graph.

So we've got noise levels have been measured, and they've been measured at 100-meter distances away from the town square.

So at 100, 200, 300 metres, and so on.

And the noise levels have been measured and then plotted onto the graph, and you can see how the noise levels change over those distances.

Right, so general trend, this is where geographers will describe the general pattern that they can see in the data.

So here we have Aisha, and she's saying the general trend in this data is that noise level decreases as one moves away from the town square.

And you can see that is exactly what is happening.

The general trend is that that line is going downwards.

Then we've got the range of data.

So this is where geographers describe the amount of data collected and the minimum and maximum values.

So here we've got Jacob, and he's talking about the range of data.

He's saying there's six plots of data, and they range from 70 to 34 decibels.

So he is looking at the top figure and the bottom figure there, and they cover an area from the town square to 500 metres away.

Perfect, so it covers an awful lot of information about the range of data.

Then we want to get a little bit more detailed.

So we're going to be looking at a quote of data here.

So some kind of individual element of the data needs to be discussed.

And as Sam says, the greatest drop in noise levels occurs between 100 metres and 200 metres away from the town square.

Let's just take a look at that.

If you look in the graph, you can see that that is the section of the graph which has the steepest bit of the line.

Between these points, the noise drops by 28 decibels.

Sam's done something really clever there.

What she's done is looked at what that top value is at 100 metres, what the value is at 200 metres, and then has taken one away from the other, just to create a really simple bit of mathematics and then a quote of the data.

Finally, the exceptions.

So geographers highlight any data that doesn't really appear to follow the general trend.

So here we've got Jun, and he says, noise levels decrease as you move away from the town square.

We've already established that's our general trend.

We're happy with that, except, he then says, at 500 metres where they rise up again slightly.

And if we look at our graph, we can see, yes, the line is generally going down, and then just at the end, it just starts to rise up again.

But Jun goes even further than that, and he says they rise again slightly from 34 decibels to 38.

So he's included a data quote in there as well.

So that's a really great bit of description from Jun.

Let's check our understanding.

A geographer has to be good at maths to successfully analyse quantitative data.

Is that true or false? Have a think.

Pause the video, and I'll come back to you in a moment.

So do geographers have to be good at maths? No.

But why is that? What's really the case here? Right, so successful analysis in geography actually means that geographers need to correctly understand what numerical values mean.

They don't necessarily have to be good at processing and manipulating the data.

When geographers analyse how one variable affects another, they describe a correlation, and there are different forms of correlation.

Let's look at those now.

First of all, a positive correlation, and this is where an increase in the value of one variable increases the value of the second variable.

You can see it in the graph there.

As river channel depth increases, so too does the velocity of the river.

When you have a line of best fit, it's worth noting that the closer the points are to that line, the stronger the correlation.

So with the graph as it is at the moment, I would say there is a correlation, definitely.

It's not as strong as it could be.

The other form of correlation is a negative correlation, and this is where the increase in the value of one variable decreases the value of the second variable.

So let's look at the graph again.

Here, we've got two different variables.

So we've got river channel depth.

And as that's increasing, we can see that the mean bedload size is decreasing.

So this is a negative correlation between those two variables.

Now, what if there's no trend line at all? Well, where there's no trend and no line of best fit, it means that there's no correlation at all between the variables.

It's a completely random idea.

So you can see two variables here, The river channel depth again and then the height of the nearest tree.

Now, we wouldn't expect there to be a correlation between those two things, so it's not really a surprise, perhaps, that there's no trend line possible on that graph.

So geographers can also manipulate data in very simple ways to create a little bit of extra meaning.

And this could mean that they find the measure of central tendency.

There's three different forms, so let's have a look at those.

In this example, Izzy has asked 10 members of the public to score the cleanliness of the town out of five.

And you can see here the 10 scores that were given.

As you might expect, some people think it's really clean, some people think it's really dirty, but most people are gonna sit in somewhere in the middle.

If we're looking for the mean value, the mean is the total of all the scores added together and divided by the number of scores.

In this example, it would be 2.

8.

So I've added together all of those scores and then divided it by 10, the 10 members of the public, and 2.

8 is my result.

Now, instead, I could do the mode value, and this is the most common value that I see in the set.

So I'm looking across my row of data there, and I can see that there are more number two scores than there are of any others.

So in this example, my mode is two.

I could also look for the median value.

This is the middle value when you put the scores in numerical order.

So I've got 10 values here, so I'm gonna count five from one end, and I'm landing on a two, and I'm counting five from the other end, and I'm landing on a three.

So I'm gonna take the middle between those two values, and that's 2.

5.

Right, let's check your understanding.

Which statement best describes this correlation? Is it a strong positive correlation, a weak positive correlation, a strong negative correlation or a weak negative correlation? You can see you've got all the options possible there.

Pause the video, have a good look at the trend line, and look at the position of the points around it, and make a decision.

Okay, so I can see that the points are actually really close to the trend line, so I know it's gonna be a strong correlation.

Is it positive or negative? Well, my line is sloping down towards the bottom right-hand corner.

So as river channel depth increases, my bedload size is decreasing.

So I know that I'm looking at a strong negative correlation.

Hope you got that one right.

Okay, our first practise task of this lesson.

Using the GRaDE acronym, write a description of this quantitative data.

So we can see we've got river depth, and we've got distance from source, and we can see how that changes, represented through a line chart.

Use the GRaDE acronym to write the description.

Pause the video.

It's gonna take a little while, so don't worry.

Pause the video and then come back to me.

Right, let's go through it in the order of the acronym.

So my general trend, as the distance from the source increases, the depth of the river increases too.

So as we would expect, the river is getting deeper the further away we get from the source.

Then we're moving on to our range of data.

There are six plots covering a range of 1.

5 metres in depth over 10 kilometres.

Maybe I could quote some data next.

The greatest depth is 1.

75 metres, and the smallest is 0.

25 metres.

And then finally my E, my exception, an exception to the trend is at two kilometres from the source, where the river gets shallower rather than deeper.

And we can see that on the graph 'cause it dips down rather than holding the general trend line.

You might not have exactly this wording, of course, but that the general things that you should be talking about.

Let's now look at how we might analyse qualitative data.

Analysing text generated from interviews and questionnaires can often mean looking for patterns of thought between the respondents, and we use a thing called coding.

Coding is one way of looking at these patterns.

Geographers create codes to represent different areas of thought or opinion.

For example, the code econ+ could be to do with the economy, the econ bit, and something positive, the plus.

An inquiry into the impact of tourism in a national park might focus on different impacts such as the negative environmental impacts or the positive social impacts for visitors.

Geographers can then take an interview transcript, and a transcript is a word-for-word account of what was said, and look for instances where the respondent discusses these impacts.

What we then do is use a letter or a number code to highlight certain areas of text.

So with this example, we can see someone is talking about a national park.

Let's just read this through.

The amount of litter I see definitely goes up in the tourism season.

Some of this gets eaten by local wildlife.

The sense of wellbeing and closeness to nature that the national park brings is undoubtedly good for people's health, but all this gets lost if there are too many people.

So there's quite a lot of different opinion going on there, and definitely, they're talking about the impact of tourism.

Let's look at the codes that we could use.

So definitely, there's a sense there that there's some negative environmental impacts that visitors bring to the national park.

So I'm gonna use the code env with a minus symbol.

We can see here the person is talking about the amount of litter, and then they're saying that it's going eaten by local wildlife.

Those are two key things which I would highlight with my env-, my n- code.

But they're also talking about how the national park might be a really good thing for people who are visiting.

So there's a positive social impact to be had as well.

And I'm going to use the code soc+, or S-O-C plus.

Here you can see they're saying there's a sense of wellbeing that comes from being in the park, and it's good for people's health.

So I would highlight those areas, and I would include my code next to them.

Let's see if you can do this for yourself now.

So complete the coding of this transcript by selecting the right code to go in the spaces.

So I've given you the two codes you're going to be using.

You're gonna be using a negative social impact, or an soc-, and a negative economic impact.

Econ- is gonna be your code there.

Your text is when they were building the new sea wall, it was bad enough.

Everyone on this road had to breathe in the dust.

I can't sell my house now because no one wants to have that ugly view of it.

So this person is talking about a new coastal defence and the problems it's brought to their life.

There are three spaces, three different bits of the statement where you need to apply a code, have another read through, look at the code and the icons be that are going to be used, and then come back to me So we can see that the transcript has been highlighted in certain areas.

First one is that everyone on the road had to breathe in the dust.

Now, that's not really an economic impact, but if it's affecting somebody's health, that's gonna be a social impact.

So our first one there is that we've got a negative social impact, our soc-.

Then I can't sell my house now.

Oh, now this is starting to sound like something that's negative and economic.

So that's gonna go in there.

Because no one wants to have that ugly view of it, So the ugly view bit is the idea that it's got a negative social impact.

It's not a very nice thing to look at.

Hope you got those right.

Now, not all data that's qualitative comes directly from responses to questions.

We also get data from articles, social media posts, adverts, blogs, and that kind of thing.

To analyse these, geographers might carry out something called discourse analysis.

This is where geographers look very deeply at how the qualitative data has been written and why it may have been written in that way.

So in discourse analysis, the actual words that are used and the actual message is not really as important as why it's been used in that particular way.

So here's a kind of typical page that you might find on social media.

So geographers could ask questions like, how are the points structured? How are they positioned on the page? What does your eye get drawn to first? What does your eye tend to ignore? How many of the resource have been produced? Is it something that's been mass marketed? Is it only for a select group of people? What conventions of language has been used? Is it quite informal? Is it more formal language? Maybe, what language itself is being used? Why have those particular words been used and not others? There's a huge choice of language that we can use.

So why would somebody decide to write in that way and not in another? Who is the intended audience? What is their age, their gender? What's their background? In what time period was this written? Is it something that was only written in the last few days, or was it written years ago and, therefore, might reflect the opinions at that time? There's lots and lots of other questions that you could ask in discourse analysis, but the important bit is that you are thinking about why it was written in a particular way rather than what was written.

So let's check your understanding of that.

Complete the sentences by finding the missing words.

Read the paragraph and then fill in the words, and go and come back to me.

Okay, let's see how you got on.

Articles, social media posts, adverts, and blogs can be analysed using discourse analysis.

This is where geographers look carefully at the resource and ask and answer questions about how and why the data was written in this way.

Hope you got those right.

We're moving on to our second set of practise tasks.

Alex is analysing a section of an interview he had with the mayor of his local town.

The topic of the interview was how migration has changed the town.

Read the interview transcript, and then write three code categories Alex could apply to it.

Then use these codes to highlight the appropriate places in the transcript.

So you're gonna go from a transcript to a completely coded document.

Let's look at Alex's interview transcript.

There have definitely been challenges and opportunities.

Unfortunately, some people do not like outsiders.

So we have worked hard to create events where the older and newer communities can come together.

We've held festivals and workshops that celebrate our town's diversity.

We have also offered advice and support to migrants who wish to set up their own businesses here.

This has been a huge success and our high street has gone from a ghost town to a busy and vibrant reflection of every corner of the world.

Yes, some people have directed antisocial behaviour towards the community, but from the majority, the message is clear: all are welcome and all are valued.

So let's just remind ourselves this is something a mayor has said about migration into this town.

Now, remember, you are trying to come up with the codes first and then go through and highlight particular areas.

This is gonna take a bit of time, so do pause the video, and have a good think, and then I'll come back to you with some of the things which I would code in this kind of transcript.

Right, let's, first of all, think about the kind of coding that Alex could apply to the transcript.

There was lots and lots of information in there, but there was certainly some discussion about the social negative changes that have taken place.

There was also some discussion about the positives in terms of the economy and in terms of the social.

So I've come up with three.

These are my three codes that I'm going to use.

Now, you might have some slightly different ones.

So long as they apply to the message that the mayor was trying to give, that's what's important.

So let's apply those to the transcript itself.

So I've looked at, first, this idea of that some people do not like outsiders, and I've given that the code soc- to indicate a negative social change.

I've applied the same code to the antisocial behaviour that's mentioned in the last sentence as well, and you can see my code has been put in there too.

Then on the more positive side, it talks about celebrating the town's diversity and the festivals and workshops.

I thought that was quite an important thing to highlight.

So I've put in an soc+ code there to indicate a positive social change.

The other big change, which I felt had to be mentioned, was the high street going from a ghost town to a busy place.

This is definitely a positive economic outcome.

So I've highlighted it there with my ec+ code.

Now, depending on what codes you've used, different parts of the transcript are going to be highlighted for you.

So just go through and have a quick check that everything you've wanted to highlight is in there.

Now, let's move on to the third and final part of our lesson today, which looks at, how does data analysis lead to good conclusions? To make a conclusion, geographers look carefully at all their analysed data, and then they try to identify four different aspects.

First of all, empirical data or empirical evidence.

These are statements that are factual and are quite hard to dispute.

We also look at interpretations.

These are statements that come from data that strongly point towards something being true.

Next is assumptions, statements the geographer assumes to be true.

And then finally, what we term false conclusions.

These are statements where there's no evidence of something being true.

So you can see from empirical evidence down to false conclusions, we're slowly losing the amount of factual information.

Empirical evidence is where we've got the most factual information and we know that something is true.

And as we get further down, we start to assume things, we are not quite sure if they're true.

And then we've got the false conclusions, which are definitely not true.

So what geographers will do, they'll try to look for these different statements within their analysis.

To make good conclusions and to answer their inquiry question accurately, geographers pay attention to the empirical evidence and the interpretations, and they try to avoid making assumptions or false conclusions.

Geographers also have to be very careful that they do not let their existing understanding of a place influence their judgement of their data.

And this is something called confirmation bias, and it's a type of thinking that we really need to avoid, but it's very easy to fall into the trap of doing it.

Let's look a little bit more the idea of confirmation bias.

At the end of their inquiry, geographers will have a collection of evidence that they've recorded, and I've kind of just tried to show this as a circle of evidence.

We've got a load of evidence that we've recorded in the field.

Hopefully, this evidence is what they were expecting.

It matches up with their prior knowledge, their understanding, everything we've already learned and everything that we know.

If we go out in the field and we find the evidence that supports that, it's really easy.

We'd like those two things to overlap really nicely, the evidence and our prior knowledge and understanding.

That means that we can write a really clear and simple conclusion because we look at the evidence, and we understand exactly what's going on.

It's kind of matching the theory that we already know.

However, the evidence may not match up with geographer's prior knowledge and understanding, and in fact, very often, it doesn't match up.

It might only be a tiny bit of the evidence that actually does what we are expecting it to do.

If you like, the world has only behaved a little bit like you were expecting when you went out and did your fieldwork.

What then happens is a poor geographer ignores all the contradictory evidence and only concludes from that middle bit, from the small amount of evidence that matches their ideas.

They tend to think that the evidence, which is overwhelmingly against what they know and understand, must be false, must be mistakes, it must have been through error, and this is known as confirmation bias, and we've gotta be really careful that we don't do this.

Which is the only conclusion that can be made from this questionnaire data? Let's take a look at this questionnaire.

To what extent do you agree with this statement? Money has to be invested in flood defences if we want our homes to be protected.

So this is a question which has been put out to members of the public, and you can see the responses that have come back there.

So most people are strongly agreeing.

We've got a bit of disagreement as well, so five people saying they disagree.

Three strongly disagree.

So a bit of a mixed bag, but most people are saying that they strongly agree with the idea that money has to be spent on flood defences if they want their homes to be protected.

Let's look now at the conclusions that you could write from this because only one of these statements really could be made from that data.

A, most people are in favour of flood defences, B, most people think the cost of the flood defences is necessary, or C, most people think flood defences are cheap.

You will notice there's a very slight difference between those three statements, but only one of them can really only be said to be true based on the data that we have in that table.

Pause the video and have a really good think about this.

You might want to discuss it with other people around you and then see what answer you come back to me with.

Okay, let's look at that first statement, first of all.

Most people are in favour of flood defences.

Well, that's not really what they're saying.

They're saying that they accept that money has to be spent on flood defences.

It doesn't necessarily mean that they love them and want them.

Then we've got this idea at the bottom, most people think flood defences are cheap.

Again, that's not really what they're saying.

They're saying that, you know, they know that money has to be spent.

It doesn't mean that they accept, that they think it's good value necessarily, that they think they're cheap.

But then we look at B, most people think the cost of the flood defences is necessary.

And yes, that does align with people agreeing with the statement.

If people are saying money has to be invested in flood defences if we want our homes to be protected, what they're saying is that, yeah, flood defence money is needed.

It's needed if we want to be safe.

So yes, the answer is B.

The type of conclusion a geographer can make depends on the type of inquiry question they ask.

Simple inquiry questions will only allow the geographer to create simple conclusions.

So if we've got an inquiry question like this, does the River Cuckmere follow the principles of the Bradshaw model? Now, in reality, the answer to that, really, is yes or no.

Does it follow it, yes or no? We might be able to say, well, some of the variables measured had results that were not like a typical river.

Therefore, the River Cuckmere does not follow the principles of the Bradshaw model.

There's not a whole lot more we could say based on the wording of our initial inquiry question.

More complex questions, though, would allow the geographer to create more complex and interesting conclusions.

So if our inquiry question were instead to be, to what extent does the River Cuckmere follow the principles of the Bradshaw model, then it gives us a little bit more scope to talk in more detail.

So as Andeep says, the variables measured at the River Cuckmere do have some characteristics of a typical river.

However, the size of the bedload shows little variation from source to mouth.

He's used an example there to show the contradictory idea.

Therefore, the River Cuckmere follows the principles of the Bradshaw model to an extent but not fully.

Can you see how by using the phrasing to what extent at the start of the inquiry question, Andeep is able to create a much more in-depth concluding statement.

So let's check our understanding now, true or false.

For geographers to create well-developed and meaningful conclusions, they need to have inquiry questions that are complex and that create interesting discussions.

Is that true or false? Yes, it's true.

But why is it true? More complex inquiry questions mean the geographer has to write in greater depth to explain the answer to the question.

So their conclusion is going to be much more well-rounded.

Well done, okay, let's look at our final practise task for this lesson.

Study this beach profile graphic that shows the gradient of the beach and the mean sediment size at different points between the sea and the cliff.

Let's take a quick look at this graphic.

You can see on the left-hand side, you've got the sea, and then on the right-hand side, there's the cliff, and you've got the beach kind of rising up between the two.

At the bottom of the table, you can see the de gradient of the four different sections of the beach in degrees.

And then at the top in millimetres, the mean sediment size is there for each of those four sections as well.

What you need to do is match the statement about the graphic that you can see on the left side here to the correct label that's on the right side.

So you've got three different types of statement there, empirical evidence, interpretation, and assumption, and you've got three statements made about the graphic, which one is which? Part two of this is then to think about, what additional data would the geographer need to collect to turn their assumption statement into an interpretation? So once you've decided which the statements is an assumption, think, okay, what extra data would I need to be able to make this an interpretation instead? Because remember, when we're trying to conclude, we like empirical evidence, and we like interpretations.

We are not so keen on assumptions and false conclusions.

This is gonna take quite a bit of thinking.

So do pause the video and go back to the graphic if you need to, and then think about which of those statements are which, and then what you would do to improve the assumption statement.

Okay, let's go back to the three statements now.

Not an easy task, so well done if you managed to get this.

First of all, the waves are pushing the larger sediment to the top of the beach.

This would have to be an assumption.

Although we can see that the larger sediment is at the top of the beach, we don't know how it got there.

It's a massive assumption to assume that it was because of wave action.

Then we've got this idea, the steepest part of the beach is nearest the cliff.

That's empirical evidence.

We could see from the diagram that the steepest area in terms of the gradient was the one that was nearest the cliff.

It was irrefutable.

And then finally, this idea and interpretation, there's a positive correlation between the beach gradient and the size of the beach sediment.

So we had four bits of gradient, four bits of sediment, and we could see that as one increased, so did the other.

So we know that that is gonna be a correlation, and we are interpreting it in that fashion.

Then we need to think about what we would do to our assumption statement to turn it into an interpretation.

What extra data would we need to collect? Remember, our assumption was about the idea that it was the waves that was pushing the sediment to the top of the beach.

So actually, it'd be a good idea to collect some wave strength data and the direction of it.

That could come from primary or secondary sources.

It might also be a good idea to look at other high-energy coastlines and look at where the largest sediment is.

Is the largest sediment at the top of the beach? Therefore, we know that the waves are strong there.

So we would think that that would be the cause of that movement.

Like I say, not an easy task, something to think about, but well done if you managed to get those.

We have covered a lot of ground in today's lesson.

Let's do a summary of what we've learned.

Quantitative data can be analysed using the GRaDE framework and by carrying out simple statistical procedures such as the measures of central tendency.

Qualitative data can be coded or subject discourse analysis, and there are different types of statement one can make in a conclusion.

And geographers need to be careful that they're making accurate points.

Well done, some of that was not easy, but you've now got a really good understanding of what it means to do good analysis and create great conclusions in your geography fieldwork.