Intro
Explore ManuScrape for free by creating an account at app.manuscrape.org and installing the cross-platform ManuScrape desktop client. ManuScrape.org is hosted in Denmark (Europe) and managed by the main technical contributors to the project.
To ensure your compliance in the EU, please collect a Data Processing Amendment from Code Collective ApS, or set up your own ManuScrape server.
Install the ManuScrape app
To get the screen capture tools, you can install the ManuScrape desktop app using the following steps.
First, download and run the latest ManuScrape installer from Github. Choose the right file for your system (.exe for Windows, .dmg for Mac).
Windows (tested on Windows 10 & 11):
Your system might prompt for security errors, as the code signing key might be invalid at the time. If you don't trust the precompiled code, you can compile it on your own machine. If the code signature is invalid, Windows 10 will open a security window, where you’ll have to click “more info” and select run anyway.
The installer will open shortly. Once it's closed simply press the “windows-key” or press start and type “manuscrape”. Click on the program to run it.
Done! ManuScrape is now running and can be found in the bottom right of your taskbar. Right click it to access the menu.
Mac OS:
Documentation is missing. Contributions are welcome !
Create account
Create a app.manuscrape.org account to sign up on app.manuscrape.org. If you host your own version of manuscrape, your technical staff will know the URL.
Create a project
You can create your first project in the desktop application, on manuscarpe.org or your own infrastructure. Just log in, and you'll be asked to create a project.
When creating a new project, you assign your project a name. There are different parameters you can set up for your project that will come up when doing your observations. These will show up every time you make a new observation.
To make a parameter you assign a label (label box) for the chosen parameter. When setting up your parameters you can choose between different types of parameters. Some parameters allow you to set up different options to choose from within the same parameter. The different parameters are:
| Type | Description |
|---|---|
| Checkbox | A boolean value that can be either true or false |
| Date | A single date value |
| Date and time | A date and time value |
| Decimal number | Supports decimal and whole numbers |
| Whole number | The parameter shows a little box where you can add a number - if you also need decimals, you can either write them or add the parameter 'decimal'. |
| Dropdown or text | This parameter allows you to create different options to choose from. You can also write a new option during your observation if none of the other's ones are suitable. But you can only select one of the given options. The options are mutually exclusive. |
| Dropdown | This parameter gives you different options to choose from. You can not freely add another option during your observation, you can only select one of the given options. The options are mutually exclusive. |
| Multiple choice or free text | Parameter creates different possible choices to select, and the possibility of 'free text' means that you can freely add something to that parameter that you have observed, but this won't pop up as a new generated option. |
| Radio buttons | You can make different options to check off. You can add as many of these as you want. BUT the boxes are mutually exclusive - meaning you can only check off 1 box. |
| Text (multi line) | You can write more than 1 line of text - ex. If you have comments for your observation |
| Text (single line) | You can only write 1 line |
Every time you add a parameter you have an option to make it “required”. This means that you can’t save and lock the observation without fulfilling the given parameter.
Collect data
Once the ManuScrape desktop app is running, you can start an observation by taking a screenshot, a scrollshot or by creating a new empty observation. Right click on the tray icon to get the different options.
This opens a new observation draft, and prompts you to input the different data and coding parameters for your current project. Simply input the meta data, images, and files. Don't forget to save save the metadata.
When you publish you observation, it will be locked for future editing, depending on your current project settings
Detailed guide for Windows
To make an observation you find a little arrow on the bottom bar on your computer (windows) that expands your tray icons. When all tray icons are visible you will see a little manuscrape logo. Right click on the logo and different options appear:
If you wish to make a screenshot observation, press 'screenshot' (Alt + N). If you have more than one project in your manuscrape, remember to choose the right project when starting your observation (choose project). When you make your screen- or scrollshot your screen turns dark and you use your mouse to mark the area of the screen that you would like to observe.
A tap opens in manuscrape and you can now fill out the parameters you've set up when you created your project. After filling out your observation parameters press 'save metadata'.
You can also edit your image if you need to anonymize them. You press 'edit image' and a little bar comes up showing different editing options.
Press 'submit and lock' and your observation will be saved. Remember to press 'save metadata' before you edit your image or it won't be saved and you have to fill them out again. You can't submit and lock your observation if you haven't filled out the required parameters in your observation.
Screenshots and scrollshots
You have two different options for recording an observation:
Screenshots work in the same way that you are already familiar with: Select an area and done!
Scrollshots are when you want to take a shot of something that does not fit onto a single screen and thus requires you to scroll. I.e., if you want to take a screenshot of all your friends on Facebook.
When taking a scrollshot, ManuScrape registers the image at a specific pace and tries to join them to one large picture afterwards. Therefore, if you scroll too fast or too slow you might experience problems with your image, e.g., that the scrollshot failed or the image is glitching. To optimize your scrollshot settings, you can regulate the amount of rows pr. crop for taking a scrollshot:
The higher the rows pr. crop, the faster you can scroll. But keep in mind that the program is taking a rolling screenshot when you scroll so be careful not to scroll too fast.
A good idea is to start by setting it at 75 rows pr. crop and scroll at a moderate pace.
IF you have connected a phone to your computer, use two fingers to ensure that you are maintaining a steady pace throughout your scrollshot.
Once you have covered the desired parts of your screen, press Alt + S to save the scrollshot.
Anonymize
Screenshots from online observations often include personal or sensitive data, which should be recorded. In the observation pop-up simply click “edit image”. This opens the image editor, where you can overlay boxes, text fields or draw on the image. Once all personal or sensitive information is removed click “overwrite image”.
This is a very important step as it ensures that you don’t store sensitive data in your project. To stay compliant over time, it is important that you set up your own clear guidelines. Optimally, you will make these guidelines transparent by uploading them into an Open Science website as OSF.IO.
These processes are complex and include an aspect of interpretation. European guidelines in relation to GDPR may be a place to start. These can be found here.
From there it may be relevant to refer to some of the academic works that relate to how and when to remove information in specific studies and fields.
Anonymizing observation images using ManuScrape
When you press 'edit image' in there is a little bar showing different symbols of possible edits and navigators. The magnifier with plus/minus allows you to zoom in and out of our image, the hand allows you to move around in your image, the circled arrow allows you to redo your action.
You can overwrite your image with these tools:
The box - Add a colored box
The T - Add text with or without solid background.
The line – allows you to create a straight line in your image, you can likewise select color (paint brush) or line width in the bar.
Remember to press ‘Overwrite image’ and wait for the bar ‘save image’ to be completed otherwise it won’t be saved !
Collaborate
Invite colleagues and peers to collaborate. Multiple researchers can submit observations on the same ManuScrape project independently. Data quality (and validity) is safeguarded by the validation rules set up in the project configuration.
You can invite collaborators by email by going to ManuScrape.org (or your own server), select the project you want to collaborate on, and invite your colleague by email.
If the invited email does not exist, it will gain access to the project once created. The email is stored as a hash sum and is not kept in cleartext. If the invited email exists on the corresponding ManuScrape server, it will gain access to the project immediately.
Export
ManuScrape only export observations that have been submitted and locked - any drafts will not be exported. This might take a while, if your dataset is large.
At the bottom of your project page on ManuScrape, you will find three buttons to export your data:
Export Spreadsheet: exports the parameters of all of your observations in an Excel-spreadsheet with the first column containing the unique ID number of each observation.
Export Images: exports all screen- and scrollshots captured with ManuScrape. The name of each image file corresponds to the unique ID numbers in the spreadsheet.
Export Uploads: exports all other files uploaded together with the observation. Again the unique ID is included as the prefix of the file name.
This might take a while, if your dataset is large. Once exported, you can manually look through your observations in Excel, view observation images or auxiliary files. To get an even better overview, import the dataset to NVivo or another qualitative data analysis software.
Import into NVivo
There are many ways to import data into NVivo depending on what your data looks like. In the following we show how we have done at MOD-Lab. In our case, each observation typically consists of three elements:
- A screenshot
- A selection of parameters
- A netnographic note saved as a docx and uploaded to the observation using the "upload file" function
In addition, some observations also contain auxiliary files such as e.g., videos, GIFs, zoomed in pictures, etc. These are uploaded in the same manner as the netnographic notes.
To import into NVivo, start by clicking all three export buttons described above, and unpack each zip-file into separate folders.
Next open NVivo, in this example we are using NVivo14:
1. Start by dragging all the image-files (from the “Export Images”-folder) into the files-tab in NVivo.
2. Select them all and click “create as case”.
3. Next, in the top-menu, select the “import”-tab, and click the “classifications”-button and then the “classification sheet”-button.
4. Select your exported spreadsheet (make sure that the ID-numbers are in the left-most column, if not, move them there), and click next.
5. Select the “case classification”, and decide if you want to leave the various check-boxes checked or unchecked depending on your project requirements. If this is the first time you import, we recommend leaving them checked. Click next.
6. On the next two pages leave the options as is, unless your project requires something else, click next.
7. Finally, click finish. All parameters should now have been attached to the relevant observations, enabling you to structure your observations in NVivo. In the left-menu under “Cases” → “Observation Id” you will see all your observations and their individual parameters.
8. Finally, drag all other files into the same “Files”-tab where you dragged your images to.
9. Now right-click a relevant file, and select “Code whole document” and select the corresponding case. This cannot be automated, unfortunately.
10. That's it! Now if you open a relevant case, you will be able to see all files and parameters related to it, giving you an easy overview of your data.
This can happen for a couple of reasons:
- The scroll is not consistent in direction. While you can collect scrollshots in one direction or the other the scrollshot will be distorted if you by mistake scroll a tiny bit up in a downwards scrollshot.
- Your field of capture includes either a static object or a video. Try to set the field of capture to only include a field that is scrolling, avoiding the top and bottom bar for example.
- The field you are capturing changes very little, and you need to turn up the sensitivity of the scrollshot. With ManuScrape running, right click on its icon, to access settings. Here you should set rows pr. crop to 75 or 100.
You should also experiment with different rhythms of scrolling. Try counting a few seconds from when your mouse/finger is at the bottom of the page to the top of the page before moving your finger/mouse back to the bottom and repeating the scroll. Rather start out a bit slow (counting to 3) and if you still experience trouble you can take up the pace. The higher the rows pr. Crop the faster you would be able to scroll.
Make sure to keep a moderate pace - don't take too long for your mouse/finger to reach the bottom of the page to scroll again - this can stop the flow and make your image look jagged, contain duplicate content or periodically freeze.
Try marking out the area you want to observe a bit higher than planned as occasionally the first row isn’t included in the scrollshot.
Remember to press save metadata when you have made your observation. If you edit your image without saving metadata, it will delete your notes and you will have to redo them.
