generated from datatrail-jhu/DataTrail_Template
-
Notifications
You must be signed in to change notification settings - Fork 2
/
06_sharing_results_14_presentations.Rmd
218 lines (106 loc) · 20.8 KB
/
06_sharing_results_14_presentations.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
```{r, include = FALSE}
ottrpal::set_knitr_image_path()
```
# How to Give a Presentation
Data Scientists are often asked to present their work. This could be in a meeting with other co-workers or at a large conference amongst colleagues with similar interests. Regardless of where you're presenting, there are general and important guidelines to consider when preparing the materials for your presentation.
In this lesson we'll cover what to consider when preparing a slide presentation, guidelines for creating good slides, and tips for presenting material well. This lesson is largely based off of our experiences and the tips from [Zach Holman](https://zachholman.com/). Links to his work directly can be found at the end of this lesson.
This lesson will be focused on how to give effective slide presentations, like what you could generate in Google Slides. These guidelines, however, will likely prove beneficial for any oral or visual presentation you may have to give as a data scientist.
### Presentation Goals
Generally, there are three goals of any presentation. A good presentation should **inspire**, **educate**, and **entertain**. Those listening to your presentation should leave the room feeling like they learned something, with the urge to try something new, and having enjoyed learning whatever it is you presented. If those three goals are accomplished, you have without a doubt delivered a stellar presentation.
### Know your Audience
In order to inspire, educate, and entertain, the first thing you must consider is your audience. For presentations, people often say you must **know your audience**. If you have a room full of analytical people (mathematicians, accountants, scientists, etc.), you should consider presenting data and numbers. Conversely, if you are presenting to a room full of creatives (artists, photographers, writers, etc.), it's likely best to ensure that the graphics you present are visually appealing. You should always **keep your audience in mind** while preparing a presentation.
### Mind your Audience
After considering your audience and designing your presentation to be most appropriate for those listening, you must always **stay within your allotted time**. Out of respect for the time of others in the room and for others who may also have information to present, this is essential. If you have 5 minutes to talk, stop at or before 5 minutes. If you have 20 minutes to speak, stop at or before 20 minutes. Your audience **will tune out** if you go over your time limit *or* you'll be told to stop by someone else in the room. It's best to avoid either of these situations and respect the time of those in your audience. Always stay within your allotted time.
### Prepare
Putting together a good presentation takes time. Be sure to leave yourself enough time to put a good story and beautiful slides together.
### Tell a Story
With those goals and your audience in mind, you're ready to start considering *how* you're going to go about presenting information to your audience. As mentioned in a previous lesson, when you're giving a presentation, you're **telling a story**.
When presenting, you're telling a story with and introduction, a middle, and an end. It's best to **outline your presentation before you start preparing slides**. This includes jotting down or at the very least thinking about what you'll convey to your audience and how you'll make it into a story. You may consider a number of questions: What story do you want to tell? What do you want to convey in your introduction? What pieces are necessary to the story to be included in the middle? What do you want your listeners to take away from your presentation? How will you include this information in your conclusion? All of these are good questions to consider when outlining your talk.
Finally, as Zach Holman explained in his [talk on talks](https://zachholman.com/talk/the-talk-on-talks/), the best talks are stories and the best stories use repetition. **It's ok to repeat important information to your audience**. Don't be afraid of repeating yourself to make a point clear to your listeners.
### Slide Design
Now that you've outlined *what* you'll talk about in your presentation, it's time to decide *how* you're going to put those ideas onto slides. There are many things to consider when designing visually-appealing slides. We'll discuss the most important in this lesson.
#### Text
A few guidelines when deciding the words you'll include on slides:
* limit the number of ideas on a single slide
* limit words on slides
* encourage listening (limit reading)
* include references on the slides directly (and at the end!)
Do not try to cram all your ideas onto a single slide. And, everything you say does *not* have to be up on the slides. You're speaking for a reason. The slides should help guide you along in telling your story, but your audience should have to listen to what you're saying to get the full picture of your story.
##### Fonts
To convey textual information to your audience, choosing a good font is important. It's generally safe to **stick to Sans Serif font**. Sans Serif fonts can be searched through at [Google Fonts](https://fonts.google.com/). And, when in doubt, **Helvetica is a safe font choice**. Further, use consistent fonts throughout your talk. More than one font *can* be used to draw viewers attention to something, but typically there should be a consistent font used throughout the presentation. Finally, as with most rules, there are exceptions, so take time deciding on what font looks best for each part of your presentation.
##### Size
Regardless of what font you choose for your text, your **text should be very large**. It's often argued that you should design your slides for the back of the room.
The same goes for images. There's no need to leave empty space on your slides around images. **Expand your images** to take up as much space on the slide as possible.
#### Colors
Finally, with regards to slide design, colors matter. Choosing a dark background (such as dark gray) means that your text should be light (maybe white). If your background is light, your text should be dark. Contrast will make it easier on your audience to read.
Additionally, use color to focus on text you want your audience to focus in on. If all your text is white, highlight a few words in a brighter color to draw your listener's attention to what's important.
The same goes for figures and plots. All images should have colors that are viewable and distinguishable by every person in your audience. Consider color blindness. Consider colors that tend not to project well (such as yellows or light pastels). Then, design your slides using colors that will work for your audience.
#### Alignment
Visually, things should be aligned. Center things in the middle of your slide. Make sure text boxes that should line up do line up. It may seem trivial, but it's important to line things up so viewers aren't distracted by something like alignment and can instead focus on your talk.
### Presenting
When it comes time to give the presentation, there are a few things to keep in mind. First and foremost, **speak clearly**. Whether you are a loud-talker, a soft-speaker, someone who speaks slowly, or a fast-talker, the most important thing is to make sure that the words you say are clear to your audience.
Second, you should **be excited**. If you're not excited or passionate about what you're talking about, why would your audience be? If you're not excited, it may be best to *reconsider your outline* and re-work your story until you are.
To avoid being boring during your talk, you can vary the speed and volume of your voice. You can move around so you're not always in one place. And, you can use pauses effectively. It's ok to not speak every second that you're up there. You can pause and give your audience some time to look at a graph before you talk about it. Silence for short periods of time is allowed!
Finally, be yourself. If you naturally like to tell jokes, tell jokes during your presentation. If stories are more your style, work a story or two in. Ultimately, if you're not comfortable in front of your audience, check to make sure you're being true to who you are as an individual.
Finally, when presenting it's important to remember that **you've seen the data and figures you're presenting many times, but your audience has not**. They'll need time to get up to speed! Thus, it's important to always introduce your axes (tell your audience what is on the x-axis and what is on the y-axis), explain the colors on your plot, *and* walk the audience through the figure you're presenting.
### Practice
Most people are hesitant when it comes to speaking. They're nervous about getting up in front of a crowd. If you find yourself being nervous to make a data science presentation, know that you're not alone.
For those who are nervous to speak publicly, nerves may never go away completely, but there are ways to *limit your nerves*. **Practicing your presentation** is one of the best ways to limit nerves during your presentation. You should present out loud to yourself. Recording yourself and listening can help too (even if it is sometimes painful)!
In addition to limiting your nerves, practicing will help you figure out the best way to explain the details included in your presentation *and* will ensure that your presentation does not take more time than you have been allotted.
The further along you are in your career, the less you will likely have to practice before presentations. However, as you get started, practicing is the best preparation for giving a talk!
#### Be Prepared
The day of your presentation, things may go wrong. Power goes out, projectors don't always present your material exactly the same way it looks on your computer screen. Technology doesn't always behave as we expect it too. In these cases, you just have to roll with the punches.
To avoid any big issues, it's best to bring a copy of your talk on a USB drive, carry any adapters you may need should you have to (or choose to) present from your own computer. Don't expect the location to have the adapter you need. And, be sure your phone is on silent. After that, just roll with anything that goes wrong. The audience will be forgiving, as they all know that technology can sometimes act unexpectedly.
### Handling Questions
After your presentation, there may be time allotted for the audience to ask questions. When answering questions, always take a little bit of time to pause and think about the question before answering. In fact, it's often helpful to repeat the question asked of you back to the audience. This ensures that the entire room has heard the question and gives you some time to think. If you need more time, remember, silence is ok! Finally, if you don't know an answer, it's ok to say "I don't know." You should never make up an answer because you're afraid of saying "I don't know."
### Example Presentation: Analysis
Often, as a data scientist, you'll be tasked with presenting an analysis you've done. To do so, you'll want to use the same guidelines previously discussed in this lesson, but with a focus on **presenting the story of your data analysis**. This includes the plots, analytical approach, and findings of your project. That said, your slides should still be well-designed, you should still tell a story, and your text and colors should all be readable when projected; however, in this case, there are a few additional caveats to consider.
To discuss these, we'll use an example presentation from Julia Silge, who is a data scientist at Stack Overflow and a wonderful presenter. In 2017 at the rstudio::conf, [Julia Silge](https://juliasilge.com/about/) presented "Text Mining the tidy Way". The [video](https://www.tidytextmining.com/tidytext.html) and [visuals](https://speakerd.s3.amazonaws.com/presentations/d6041c6b704d4bf7bfa9d0973ae2d006/rstudio_conf.pdf) from her presentation are available online. This talk discusses the `tidytext` package (which she and [David Robinson](http://varianceexplained.org/about/) developed) and example analyses using this awesome R package. We've discussed this package previously, but if you want a refresher on tidytext, how to work with textual data, and sentiment analysis, [this](https://www.tidytextmining.com/tidytext.html) is a great resource!
As above, we'll highlight a few slides from the talk and discuss considerations to make when presenting a data analysis.
#### Talk Outline
As in Lucy's talk, Julia Silge, in addition to an introduction and a conclusion, had three main parts to her talk:
* `tidytext` package
* Jane Austen example
* NASA example
In data analysis talks, it's often best to start with background information. Julia covers this information when discussing what the `tidytext` package is. Then, she delves into discussing a few examples of the types of analyses you can do with this package.
#### Slide Design
As discussed above with Lucy's presentation, in Julia's slides, she limits the amount of information she puts on any one slide, thus allowing the audience to listen to the words she's saying. And, she has text that's large enough and legible when projected.
However, what's different about this presentation is how Julia presents her analyses. When presenting an analysis, it's generally best to first explain the approach and then present the results. For example, Julia uses TF-IDF to analyze text in Jane Austen's novels.
She first explains what TF-IDF is. She includes the bare minimum amount of information on her slide, so that her audience is only seeing what's necessary. Then, she verbally describes what TF-IDF is.
On subsequent slides she includes the code she used to run the analysis, discussing what is going on with the code on each slide and describing the output.
Finally, for each part of her analysis, she includes a visualization of the results. These figures have text large enough to be seen by all in the audience, are well-labeled, and include an informative title.
Note that Julia does not go into every single detail of every analysis, nor just she talk about all the things she tried in a single presentation. She includes enough information so that her audience learns and can go try this type of analysis on their own, if they're interested!
#### Presentation
Watching the [18 minute video](https://www.tidytextmining.com/tidytext.html) will give you the best sense of how Julia presented the information in her `tidytext` presentation; however, we'll highlight a few things here.
Julia spoke at a reasonable pace and clearly, keeping her audience interested. Additionally, she presented true to her personality. Julia and Lucy are two different people and that is clear in their presentation styles. Differences are wonderful and help keep presentations interesting! It would be boring if everyone were the same.
Additionally, Julia knew her audience. She knew that there would be many programmers in her audience, and that most of them would be familiar with the R programming language. Thus, she didn't need to explain what the R syntax was. She could put R code on her slides, and her audience would successfully follow along. But, she knew that not everyone has worked with the `tidytext` package. Thus, she focused her time here, discussing the purpose of and basics on how to use the `tidytext` package and provided clear examples of analyses using this package. When outlining and preparing presentations, always consider your audience. If your audience is more technical, include more details about *how* you did the analysis. If they're less technical, be sure to explain what you did at an appropriate level. And, of course, always make sure your figures are clear, regardless of your audience.
Additionally, we'll note that at around 17:45 in the presentation, Julia went to show an image but it didn't immediately show up how she had intended. This is a great example of rolling with the punches. She didn't panic. She simply acknowledged that it didn't work the first time and tried again. The audience, of course, didn't mind, but when you're presenting, these small occurrences can feel like a big deal. When it happens to you in the future, know that things don't always work as anticipated and that this happens to everyone. If this happens to you, take a play out of Julia's book and simply acknowledge what happened, try again, and continue on!
### Summary
In this lesson we've discussed the important pieces to designing and presenting your data analysis project. Always be mindful of your audience, create good slides, and present your ideas clearly. Practicing your presentations will help, especially when you're first giving talks.
### Additional Resources
#### Presentation Guidelines
* [The Talk on Talks](https://zachholman.com/talk/the-talk-on-talks/), by [Zach Holman](https://zachholman.com/)
* [speaking.io](https://speaking.io/), Thoughts on Public Speaking from [Zach Holman](https://zachholman.com/)
* [VIDEO: Your (first) talk](https://www.youtube.com/watch?v=4UxTgWkFJME), by [Steph Locke](https://itsalocke.com/)
* [How Typography Determines Readability: Serif vs. Sans Serif, and How To Combine Fonts](https://medium.freecodecamp.org/how-typography-determines-readability-serif-vs-sans-serif-and-how-to-combine-fonts-629a51ad8cce), by [Harshita Arora](https://twitter.com/aroraharshita33)
* [Google Fonts](https://fonts.google.com/)
#### Presentation Examples
* Text Mining the tidy Way [video](https://www.tidytextmining.com/tidytext.html) [visual](https://speakerd.s3.amazonaws.com/presentations/d6041c6b704d4bf7bfa9d0973ae2d006/rstudio_conf.pdf), from [Julia Silge](https://juliasilge.com/about/)