connecting convo sarah paul james 01 — Transcript

Full transcript

1. 0:00thank you for joining us for the
2. 0:02disinformation
3. 0:03misinformation and fake news teach out
4. 0:05i'm julia one of your mentors
5. 0:07i'm a graduate fellow for teachouts at
6. 0:09the center for academic innovation here
7. 0:11at the university of michigan
8. 0:12and i'm a grad student that studies
9. 0:14learning design and information science
10. 0:16i'm here with sarah brandt james park
11. 0:18and paul resnick three of our teachout
12. 0:20experts and we're going to talk you
13. 0:21through some of the connections that you
14. 0:23learners have made
15. 0:24between the two videos that james and
16. 0:26sarah created with the three of you mind
17. 0:28reintroducing yourselves for our
18. 0:30learners
19. 0:32sure thing i'll start uh hi my name is
20. 0:34sarah brandt
21. 0:35i work at newsguard where i head up our
22. 0:37news literacy programs
23. 0:39and newsguard is a service run by
24. 0:42journalists that
25. 0:44rates the credibility of all these
26. 0:46websites
27. 0:47hi i'm james park i'm the assistant
28. 0:50director of the center for social media
29. 0:52responsibility here at the university of
30. 0:53michigan
31. 0:54and i work really closely with paul
32. 0:56reznick who you're about to meet
33. 0:59hi i'm paul resnick i'm a professor at
34. 1:02the school of information and associate
35. 1:04dean for research and i'm the
36. 1:05director of the center for social media
37. 1:07responsibility
38. 1:09excellent thank you all so much for
39. 1:11being here um so i'm going to go ahead
40. 1:13and start with some questions that i've
41. 1:15pulled together
42. 1:16um based on some of the connections that
43. 1:18our learners made
44. 1:19in the discussion forums so our first
45. 1:22question
46. 1:23is that after watching the videos that
47. 1:25james and sarah contributed to the teach
48. 1:26out
49. 1:27learners had a lot of different
50. 1:28perspectives and questions surrounding
51. 1:30the two purposes and goals of your sites
52. 1:33so what do you think the differences are
53. 1:35between newsguard and iffy quotient
54. 1:38and what are the overlaps um sarah let's
55. 1:40start with you
56. 1:42sure so i'll start with the overlaps i
57. 1:44think
58. 1:45what newsguard and if you quotient are
59. 1:47both
60. 1:48designed to do is contribute
61. 1:52to this large effort of addressing
62. 1:54misinformation
63. 1:55online particularly on social media
64. 1:59and we go about that goal in different
65. 2:01ways
66. 2:02so at newsguard what we do is we rate
67. 2:05the credibility of news sources and we
68. 2:08produce reviews of those news sources
69. 2:11and the idea is that
70. 2:13we want to empower information consumers
71. 2:16to learn more about the the news
72. 2:18articles and the sources they're seeing
73. 2:20on social media and on google so that
74. 2:22they can make better informed decisions
75. 2:23and maybe not be duped by misinformation
76. 2:26or share with their friends
77. 2:29whereas the iffy quotient is
78. 2:33using as an input what
79. 2:36what sarah's site is is producing
80. 2:39so they're evaluating all of the sites
81. 2:44and their original intent was for end
82. 2:46users
83. 2:47consumers to use this to make better
84. 2:49choices about what they're reading
85. 2:51we're using it to try to make an
86. 2:52aggregate measure
87. 2:54of how well the platforms are doing so
88. 2:57we take
89. 2:58their ratings of the sites we we get all
90. 3:00of the
91. 3:01you know the which urls are popular each
92. 3:04day
93. 3:04on facebook and twitter and we say how
94. 3:06many of them
95. 3:07are news guard red versus how many of
96. 3:09them are new scarred green
97. 3:13james do you have anything to add i
98. 3:16think
99. 3:16both sarah and paul have have really
100. 3:18covered it really well the only thing i
101. 3:19would add i think
102. 3:20is just to remind everyone that um
103. 3:24newsguards methodology and our
104. 3:26methodology
105. 3:28uh are kind of put together in a pretty
106. 3:31clear
107. 3:32and organized fashion so that it's
108. 3:34applied
109. 3:35kind of uniformly to everything so
110. 3:38you know it's a transparent set of
111. 3:40criteria used
112. 3:41and we're not trying to kind of play
113. 3:44favorites or anything everything gets
114. 3:45treated
115. 3:46the same way and it's the same case with
116. 3:47uh facebook and twitter are measured the
117. 3:49same way
118. 3:50so we're trying to be kind of as a as a
119. 3:53uniform and as objective as possible
120. 3:56awesome that's actually a really great
121. 3:58segue to my next question
122. 4:00so one thing that's come up not
123. 4:03particularly
124. 4:04in the connecting conversation space but
125. 4:06kind of throughout the teach out as a
126. 4:08whole
127. 4:08is that learners are really curious
128. 4:11about how
129. 4:12experts in the field of miss or
130. 4:13disinformation
131. 4:15such as yourselves keep personal biases
132. 4:18from trickling into your work
133. 4:20so james you kind of alluded to that a
134. 4:22minute ago but i'd love to hear
135. 4:24a little bit more in depth from the
136. 4:25three of you how do your experiences
137. 4:27with
138. 4:28creating and maintaining tools like iffy
139. 4:30quotient and newsguard
140. 4:32how do you keep your personal biases out
141. 4:34of that process
142. 4:37uh paul maybe we'll start with you this
143. 4:38time okay
144. 4:40um so uh i think
145. 4:43james has really articulated the
146. 4:46the key is that we need to have um
147. 4:50we need to have uniform criteria that
148. 4:52are that are spelled out
149. 4:53and and we have transparency so that
150. 4:56people can hold us accountable we're
151. 4:57sort of in the business of
152. 4:59holding uh twitter and facebook
153. 5:01accountable
154. 5:03but uh we need to be accountable as well
155. 5:05so we've
156. 5:07we've uh you know put out
157. 5:10uh you can click through to find all the
158. 5:13details of what our methods are
159. 5:15and and uh you know newsguard is not the
160. 5:19only place out there that is that is
161. 5:23rating websites we've chosen them as our
162. 5:26as our primary source for the for the
163. 5:28ratings
164. 5:29precisely because they are the most
165. 5:32transparent about what their process
166. 5:34is for rating for raiding the news sites
167. 5:38so i'd say transparency is is our
168. 5:40biggest
169. 5:42um way of holding ourselves and allowing
170. 5:45the public to hold us accountable
171. 5:47for those decisions that we're making
172. 5:49where personal biases could come in
173. 5:52and i appreciate that in the bios you
174. 5:54know i was going through the
175. 5:56um i'm part you know i just finished my
176. 5:58school of information degree but i
177. 6:00appreciated that but you know when i
178. 6:02went to look at the
179. 6:03center for social media responsibility
180. 6:05site for the first time
181. 6:07that you all have kind of your like
182. 6:09professional affiliations and your
183. 6:11interests
184. 6:11and you know your funding sources kind
185. 6:13of in your bios so it's very clear you
186. 6:16know
187. 6:16on the site what exactly um you know
188. 6:19where exactly your perspectives are
189. 6:21coming from which i think
190. 6:22um learners really appreciate yeah
191. 6:25sarah what about you from the newsguard
192. 6:27perspective yeah i think
193. 6:29um would love to echo a lot of what paul
194. 6:31had to say
195. 6:32i think transparency is paramount and
196. 6:35we've tried to build as much
197. 6:36transparency into our process at
198. 6:38newsguard
199. 6:38as well so likewise we have all of the
200. 6:41names and biographies and contact
201. 6:43information
202. 6:44for every single one of our journalists
203. 6:46analysts or editors
204. 6:47you know we name and identify all of our
205. 6:50major funders
206. 6:52and we use a
207. 6:55very transparent process uh basically
208. 6:58what we do is we write every single
209. 7:00website
210. 7:01using the same nine criteria that
211. 7:04evaluate both their credibility and also
212. 7:06the transparency of the website so we're
213. 7:08actually looking at how transparent are
214. 7:09these websites
215. 7:10as well um you know we've tried to build
216. 7:13in a few different
217. 7:14safeguards against bias into our work
218. 7:18as well in addition to you know applying
219. 7:20those nine criteria
220. 7:21consistently we make sure that
221. 7:24every single one of our reviews
222. 7:28is seen and edited by multiple people on
223. 7:31staff
224. 7:32and some of the reviews are actually
225. 7:34discussed by our entire staff
226. 7:37and every single review we ever publish
227. 7:39is edited and discussed
228. 7:40and reviewed by our two co-founders who
229. 7:44serve as a kind of uh great balance at
230. 7:46the top and
231. 7:47and one thing i really appreciate but
232. 7:49appreciate about the two of them
233. 7:50uh stephen brill and gordon grovitz is
234. 7:52that they're actually kind of on
235. 7:53opposite sides of the political spectrum
236. 7:55personally so steve is traditionally a
237. 7:57bit more left-leaning
238. 7:59in his views whereas gordon is a bit
239. 8:00more right meeting his views so i think
240. 8:02having those two review every single one
241. 8:04of our
242. 8:04reviews is really helpful to you know
243. 8:06keep us honest and fair
244. 8:08in our review process um and you know
245. 8:11what one thing i
246. 8:12i do want to also acknowledge is the
247. 8:14fact that you know
248. 8:16ultimately um we're humans uh
249. 8:20we're humans producing these reviews and
250. 8:22you know james and paul and his team are
251. 8:24humans designing the iffy quotient
252. 8:26and anything that's done by a human is
253. 8:28inherently going to be influenced to
254. 8:30some extent by bias
255. 8:31you know what we can do is try to
256. 8:33minimize the impact of that bias and
257. 8:36and just be aware of our own biases and
258. 8:39acknowledge that they do exist and
259. 8:41understand how they're influencing what
260. 8:42we do and how we consume information
261. 8:46definitely james any thoughts
262. 8:49i again paul and sarah have had a lot of
263. 8:52the great
264. 8:53points um the one thing i will add here
265. 8:55in addition to
266. 8:56sarah's final comment is yeah we're all
267. 8:59human and we're not perfect
268. 9:01but i think that on top of that um the
269. 9:04great thing about humans over time is
270. 9:07that we've also developed a lot of best
271. 9:08practices and we've developed
272. 9:10methods that various fields as a whole
273. 9:14have kind of settled on
274. 9:15are good methods and best practices for
275. 9:18that particular field so in the case of
276. 9:20newsguard
277. 9:20or in the case of us with iffy quotient
278. 9:23we're adhering to a lot of the
279. 9:25best professional practices of our
280. 9:27fields so as to kind of
281. 9:29um try to crowd out as much of the bias
282. 9:32or
283. 9:32or human fallibility as we can you know
284. 9:35we've developed these kinds of systems
285. 9:38as large groups of people to try to you
286. 9:40know
287. 9:41maintain objectivity where we can
288. 9:44and to settle on what are the better
289. 9:47things for us to do
290. 9:48as a a field so i think that's what we
291. 9:51try to lean on
292. 9:52in our all of our work and i'll just say
293. 9:55that you know
294. 9:55for the iffy quotient um
295. 9:59there's a new number produced every day
296. 10:01what's the what's the fraction of the
297. 10:04of the popular articles on facebook that
298. 10:06were
299. 10:07from iffy sites what's the fraction from
300. 10:09twitter that were those numbers are
301. 10:10produced every day
302. 10:12no human hands touch those numbers
303. 10:14that's all
304. 10:15automated scripts that are running and
305. 10:18uh
306. 10:19if we change something we're not
307. 10:21changing it for one day we're
308. 10:23changing a calculation uh and we're
309. 10:25changing the algorithm
310. 10:27that will run every day for a long time
311. 10:30so uh it's not quite so easy to cook the
312. 10:33books that way
313. 10:35because because you're not you can't go
314. 10:37in and just manually adjust
315. 10:39individual numbers
316. 10:44so a lot of learners um this is kind of
317. 10:46a
318. 10:47segue but um a lot of learners said that
319. 10:50they
320. 10:50are being exposed to newsguard and iffy
321. 10:53quotient and other tools like this
322. 10:55um for the first time because of the
323. 10:57teach out
324. 10:58which is really exciting we have a lot
325. 10:59of international learners learns from
326. 11:01outside the united states where
327. 11:03u of m is um that are
328. 11:06kind of discovering these tools for the
329. 11:08first time so i'm interested to hear
330. 11:10about
331. 11:11how um or what advice you three would
332. 11:14have
333. 11:14for people that are starting to use
334. 11:16these tools to become more
335. 11:17news and media literate um what advice
336. 11:20do you have
337. 11:21uh james let's start with you this time
338. 11:24oh boy um i think
339. 11:28i'll start by saying that i think a good
340. 11:30practice in all of this
341. 11:32is to use a healthy dose of common sense
342. 11:35along with um any tools that you have at
343. 11:38your disposal
344. 11:39so in the realm of social media we find
345. 11:41that it's really easy
346. 11:43to have a knee-jerk reaction where you
347. 11:46just click
348. 11:47and share something because you read it
349. 11:49and it
350. 11:50immediately kind of sparks something in
351. 11:52your brain but i think that
352. 11:53one of the things that i hope that
353. 11:56someone could
354. 11:57take away with iffy quotient and
355. 12:00related tools would be if you see
356. 12:02something that
357. 12:04sounds too good to be true or
358. 12:07if you happen to look at a details page
359. 12:09on the input quotient and you see
360. 12:11a set of urls that you've come across
361. 12:14before or a set of urls from a site that
362. 12:17you've
363. 12:17seen before or one you haven't also
364. 12:20think about
365. 12:21what are the implications of sharing
366. 12:22that information with other people
367. 12:24if you don't have the full confidence
368. 12:27that
369. 12:28it's being reported in a objective and
370. 12:31fair way and it's kind of the same way
371. 12:33when you tell someone just any
372. 12:35old thing that you heard from the rumor
373. 12:37mill um
374. 12:38and if you did then you kind of also
375. 12:40should understand what the implications
376. 12:42are you know someone said something
377. 12:43untrue or questionable about your
378. 12:46brother and then just went ahead and
379. 12:47said it to anyone they came across on
380. 12:48the street
381. 12:49that has implications um and if there
382. 12:52were an iffy quotient for information
383. 12:54about your brother then you would hope
384. 12:55that someone would use that tool before
385. 12:56spreading rumors
386. 12:57so
387. 13:01sarah what do you think yeah um well i
388. 13:05guess
389. 13:06i want to answer in two parts so
390. 13:09advice for using useful specifically but
391. 13:11also just advice for using
392. 13:13any kind of tool that helps you avoid
393. 13:15misinformation
394. 13:16more generally in terms of using
395. 13:19newsguard for the first time
396. 13:21so the great thing about newsguard is um
397. 13:24it's really easy to once you have it
398. 13:28installed see at a glance
399. 13:29uh the red and green reading of a source
400. 13:31and
401. 13:32just get a really quick indication of
402. 13:34the source's reliability
403. 13:36but what i would particularly recommend
404. 13:39especially for first-time users
405. 13:40is to go beyond the red and green rating
406. 13:42and actually click
407. 13:43and open it and read the full nutrition
408. 13:45label review because
409. 13:47you'll get a much more nuanced
410. 13:48understanding of the new sources
411. 13:50that you're encountering that way um
412. 13:53it's even a useful exercise for
413. 13:54news sources that you've been consuming
414. 13:56for years and you think you're familiar
415. 13:58with
416. 13:58you might still learn quite a bit from
417. 14:01reading your label
418. 14:02but then in terms of just advice more
419. 14:05generally speaking
420. 14:06um you know i think what's really
421. 14:09important is to acknowledge that
422. 14:10a tool is just a tool it serves a
423. 14:14specific purpose but it shouldn't be
424. 14:15seen as like a panacea or
425. 14:18used as a crutch um so it can be useful
426. 14:21in certain circumstances but
427. 14:23um it's only effective if you're also
428. 14:25applying just critical thinking and as
429. 14:28james said
430. 14:28common sense and also try to think about
431. 14:32ways that
432. 14:33you can apply uh what you've learned
433. 14:35from using that tool to other
434. 14:36other walks of life so like with
435. 14:39newsguard for example
436. 14:40we rate news websites but
437. 14:43you could think about the way that we
438. 14:45approach reading these websites to to
439. 14:47try to
440. 14:48use those same criteria yourself when
441. 14:50you're assessing things that
442. 14:52maybe you wouldn't traditionally rate
443. 14:54like the credibility of a youtube
444. 14:55account or something like that
445. 14:58i think that's really helpful for for a
446. 15:00lot of people kind of to expand the tool
447. 15:02outside of um
448. 15:05or expand the uh the uh
449. 15:09use of the tool you know the the
450. 15:11structure of the tool to context
451. 15:13outside of like where the tool is used i
452. 15:15think that's like really helpful
453. 15:17um so next question um this is
454. 15:20specifically for james and paul
455. 15:22over some things that came up about iffy
456. 15:24quotient um
457. 15:26but sarah i'd also be interested you
458. 15:27know because you work so closely with
459. 15:29them
460. 15:29um whether or not you have some thoughts
461. 15:31as well so
462. 15:32um learners were really interested in
463. 15:34how something like iffy quotient
464. 15:36um you know which you know gives a
465. 15:38rating to
466. 15:40platforms like facebook and um twitter
467. 15:43um could be expanded to detail quotients
468. 15:46beyond those media social media sites
469. 15:48say for instance um tracking the
470. 15:50iffyness
471. 15:51quote unquote of a single news source
472. 15:53article over time
473. 15:56or single news sources articles over
474. 15:58time excuse me
475. 15:59um is this something that you've thought
476. 16:00about for the tool or is this something
477. 16:02that you don't really see
478. 16:04being within the tools per view
479. 16:08well yes it's a great question the uh
480. 16:12we are well newsguard is writing sites
481. 16:15not articles and
482. 16:19uh we are tracking articles
483. 16:22that get popular on facebook and twitter
484. 16:25but we're not getting any information
485. 16:27about those articles
486. 16:29other than what site they came from and
487. 16:31we go and see
488. 16:32what a newscar say about the site even
489. 16:35their the iffiest
490. 16:37site the ones that get the worst ratings
491. 16:40on newsguard some of the things that
492. 16:42they publish are true
493. 16:45uh but and
494. 16:49so we don't you know when we come out
495. 16:51with something that says
496. 16:53uh 11 of the most popular urls on
497. 16:55facebook were
498. 16:57were from were iffy yesterday that
499. 16:59doesn't mean that 11
500. 17:00of the news that was shared on facebook
501. 17:02was wrong
502. 17:04there could be false information that
503. 17:07came
504. 17:08from good sites occasionally they do
505. 17:10make mistakes
506. 17:11and even the the the iffy sites
507. 17:14sometimes
508. 17:15uh they put out things that are true so
509. 17:18we would love
510. 17:20to have a better iffy quotient that
511. 17:23actually
512. 17:24measured what percentage of the news
513. 17:26that was shared yesterday on
514. 17:27facebook was true that's
515. 17:30a lot harder and we don't yet have the
516. 17:32capability to do it
517. 17:34but we do have aspirations uh and i'll
518. 17:37share in a minute
519. 17:38what that might look like um and if we
520. 17:41were if we did have some way to actually
521. 17:43evaluate the articles
522. 17:45then we could do something like what
523. 17:47people are asking for
524. 17:50instead of you know just rating the
525. 17:53sites by their journalistic practices
526. 17:55maybe we could actually have something
527. 17:56that says well this is a site that
528. 17:58you know is ninety percent accurate and
529. 18:01this one this other site is only seventy
530. 18:02percent accurate
531. 18:04um so we don't have the capability to do
532. 18:06that today because we don't
533. 18:07have any way to we're not even trying
534. 18:10right now to evaluate individual
535. 18:12articles
536. 18:14i would like to get to something that
537. 18:15would evaluate in
538. 18:17individual articles
539. 18:20through some kind of public jury system
540. 18:23where we would
541. 18:24ask people to to write individual
542. 18:27articles ask them to do some research
543. 18:29um see if you can find other things of
544. 18:32course
545. 18:33journalists are are trained to do fact
546. 18:35checking
547. 18:36the public is not but we've been doing
548. 18:40some
549. 18:40some tests and and it appears that
550. 18:44several if you sort of take the average
551. 18:47of several opinions
552. 18:48they can do pretty well at assessing
553. 18:54whether whether stories are true false
554. 18:57misleading
555. 18:59and so we're thinking about maybe we
556. 19:00could actually run some kind of
557. 19:02um volunteer jury service where people
558. 19:06could
559. 19:06could rate could rate things and we're
560. 19:10we're trying to think about how can you
561. 19:11even maybe make it into a fun game to
562. 19:13participate in where
563. 19:14see if you can guess what everybody else
564. 19:16will say about this article
565. 19:18and uh we're hoping that by turning it
566. 19:20into that kind of game it'll make it fun
567. 19:21that
568. 19:22that people will want to participate and
569. 19:24then we'll
570. 19:25we'll get their uh aggregated opinions
571. 19:28and and
572. 19:28be able to use that maybe you know right
573. 19:31now we're tracking the the 5 000 most
574. 19:33popular urls every day on facebook and
575. 19:35twitter
576. 19:36we probably can't get that big of a jury
577. 19:38to participate
578. 19:39um that we would get you know 10 people
579. 19:42to rate
580. 19:43each of 5 000 articles but maybe we
581. 19:45could get the most popular ones each day
582. 19:47and actually get
583. 19:48assessments of um of how good the
584. 19:51individual articles were
585. 19:53sort of like a crowd-funded
586. 19:56news assessment article assessment
587. 19:59service
588. 20:00or like i said it's almost citizen
589. 20:01science in a way it sounds yeah
590. 20:03that's really cool yeah but not funny
591. 20:06you'd sort of be crowd labored yeah
592. 20:10uh crowd source maybe is the word i was
593. 20:12thinking of yeah
594. 20:14uh james do you have any additional
595. 20:17thoughts from the iffy quotient
596. 20:18perspective
597. 20:19yeah i mean i think i'm what i'm going
598. 20:21to say is probably going to tease sarah
599. 20:23up
600. 20:23but i think that one of the important
601. 20:26things
602. 20:26to keep in mind with every quotient is
603. 20:29because of our partnership with
604. 20:30newsguard
605. 20:31we're putting some of the trust and
606. 20:33responsibility in newsguard's hands
607. 20:36to continually check on the integrity of
608. 20:40various news sources that are out there
609. 20:43and maybe this is the time to turn the
610. 20:47stadium over to syracuse and she'll talk
611. 20:49a little bit more about that process
612. 20:52yeah i mean i think um
613. 20:55one thing we really wanted to make sure
614. 20:56we did that news card was not just
615. 20:59you know one and done we write a website
616. 21:00once and then we can never
617. 21:02change um so we're constantly updating
618. 21:04our reviews to reflect
619. 21:06um ways in which a site's content
620. 21:09and its iffyness might change over time
621. 21:12um and so so we do see that happen you
622. 21:15know we do see news websites these
623. 21:17practices change over time either the
624. 21:19better or worse
625. 21:20and that'll be reflected in the way that
626. 21:22our review is updated
627. 21:24um but of course that process um
628. 21:27you know isn't on the day by day article
629. 21:29by article
630. 21:30level um it is still quite aggregated
631. 21:34um you know the beauty of what we do
632. 21:37by by doing it in an entirely human
633. 21:40driven way is that it's
634. 21:42very rigorous and transparent and
635. 21:44detailed
636. 21:45but you know the downfall is that we
637. 21:48can't get to the scale of
638. 21:49reading hundreds of thousands of
639. 21:52articles um
640. 21:53and i i don't think i mean without i
641. 21:57guess crowdsourcing is one way to get
642. 21:59about that
643. 22:00uh to get towards that but in other ways
644. 22:02it's if you feel like
645. 22:04ai assessments can can get to a point
646. 22:06where they're accurate enough
647. 22:08to service some sort of proxy for
648. 22:11article reliability so
649. 22:14one really prevalent theme in the
650. 22:16discussion forums
651. 22:18um again this is something that kind of
652. 22:20goes beyond the bounds of the connecting
653. 22:22conversations but i think it's going to
654. 22:24be really important for learners to hear
655. 22:25about
656. 22:27is that learners were really focused on
657. 22:31finding the source quote unquote of a
658. 22:34piece of information
659. 22:35a piece of disinformation so um
660. 22:38you know going back and trying to trace
661. 22:40back to the the source of a
662. 22:42um you know a social media generated uh
663. 22:45you know piece of information that kind
664. 22:47of is taken on a life of its own
665. 22:49as we see happen um so so what level of
666. 22:53importance
667. 22:54would the three of you grant to trying
668. 22:56to find
669. 22:57that um that source or that kind of like
670. 23:01uh you know point where the
671. 23:03disinformation
672. 23:04sort of started um when you're on social
673. 23:07media
674. 23:08do you think that that's necessarily
675. 23:09important when you're trying to
676. 23:11discern whether or not something is uh
677. 23:14you know a good
678. 23:15uh a good piece of information to to uh
679. 23:18you know consider or is it something
680. 23:20that maybe is not quite as
681. 23:22important um paul maybe let's start with
682. 23:25you
683. 23:27well there yeah there are multiple
684. 23:29pieces of it one is
685. 23:32if you're trying to assess do you want
686. 23:34to try to figure out
687. 23:36where to come from because that might
688. 23:38make you more or less skeptical of it
689. 23:42but the other piece is if if there is
690. 23:46misinformation or just information
691. 23:48that's going around
692. 23:49can we figure out where to where did it
693. 23:51originate
694. 23:52is this russian propaganda or is it
695. 23:55homegrown
696. 23:56is it uh you know did
697. 23:59did someone did this did this originate
698. 24:02from the opposing political campaign
699. 24:05or did it originate from from some
700. 24:07independent actor
701. 24:09and i guess
702. 24:13it makes sense that people are trying to
703. 24:14do that as a as a
704. 24:16as a form of accountability um it's not
705. 24:19where we're focused
706. 24:20we're focused not on the source problem
707. 24:23but on the distribution problem
708. 24:25and so that's we're trying to create
709. 24:27accountability for the
710. 24:29the platforms which are the source of
711. 24:32their art
712. 24:33are the mechanism of widespread
713. 24:35distribution
714. 24:36and we think they have certain
715. 24:37responsibilities to
716. 24:40uh for some kinds of things
717. 24:43they should limit the reach that it has
718. 24:46and so so that's just sort of a
719. 24:49difference of focus that we're
720. 24:51we're focused on on the which things
721. 24:54become popular
722. 24:55and what are the mechanisms we can use
723. 24:58to try to keep them from becoming too
724. 24:59popular
725. 25:00as opposed to preventing anybody from
726. 25:04ever saying anything that's false
727. 25:06i think it's just not that harmful if
728. 25:07people are saying things that are false
729. 25:09if they don't get very wide distribution
730. 25:12so we're focused on the
731. 25:14which things become very popular good
732. 25:17point
733. 25:17sarah do you have anything to add yeah
734. 25:21so
735. 25:21in terms of um you know identifying the
736. 25:25provenance
737. 25:25of a piece of misinformation to the very
738. 25:28first
739. 25:29point that it originated that's
740. 25:32something we've
741. 25:33been trying to do a bit more recently at
742. 25:35newsguard um
743. 25:36i think it's a really useful exercise
744. 25:39for understanding
745. 25:40the entire life span of misinformation
746. 25:43all the way from its
747. 25:44very first point of origin to how it
748. 25:47snowballs and spreads and then
749. 25:49eventually goes viral
750. 25:51um understanding that ecosystem is
751. 25:52really interesting and i think it can be
752. 25:54identifying for people who are trying to
753. 25:57better understand encounter
754. 25:58disinformation
755. 25:59um in terms of you know if you're just a
756. 26:02general
757. 26:03information consumer trying to
758. 26:06understand whether something is
759. 26:08true or false um it might not be
760. 26:11necessary to go through that
761. 26:12pretty intensive process of tracing a
762. 26:15myth all the way back to its origin
763. 26:17um in some cases it requires a lot of
764. 26:19research and reporting
765. 26:21um you know instead what might be more
766. 26:23useful is just
767. 26:25you know checking that claim with
768. 26:27multiple different
769. 26:28credible sources um and
770. 26:32if you found multiple credible sources
771. 26:34that are relying on
772. 26:35um other you know first-hand
773. 26:38or other rigorous sourcing uh to make
774. 26:41their claims then you can feel
775. 26:43you know pretty confident that what
776. 26:45you're reading is accurate
777. 26:46or not
778. 26:50james do you have any thoughts
779. 26:54yeah this is a tough one i mean uh paul
780. 26:58and sarah have definitely
781. 27:00covered a lot of good ground um i don't
782. 27:03know that i have anything
783. 27:04kind of major to add i think
784. 27:08maybe one of the questions when we say
785. 27:10source
786. 27:13it really can come down to
787. 27:16the fact that as both paul and sarah
788. 27:19have alluded to there
789. 27:20is a point of origin for a
790. 27:24a claim or an argument or a story um
791. 27:28and that could be like just a single
792. 27:30person
793. 27:31saying it or of course it could be like
794. 27:33an event that was witnessed and so then
795. 27:35you have a lot of accounts
796. 27:36um so that could be like this the source
797. 27:39um but of course then you also have
798. 27:41a particular story you may break thanks
799. 27:44to like one
800. 27:45publication um being the first to talk
801. 27:48about it
802. 27:49and in that case then it also becomes a
803. 27:52question of
804. 27:53the trustworthiness of
805. 27:56that reporting source right
806. 28:00and of course for us um involved with
807. 28:02the proportion and for newsguard
808. 28:05uh the you know what we deal with most
809. 28:08often
810. 28:08are the kind of those reporting sources
811. 28:11and of course those reporting sources
812. 28:12often have to work you need the
813. 28:15investigative methods and what
814. 28:17not what not to um identify or even be
815. 28:20in touch with the kind of
816. 28:22actual point of origin so
817. 28:26it's complicated i would say that again
818. 28:28it's about
819. 28:30trying to develop a framework of
820. 28:34understanding and and trusting
821. 28:38reporting sources that tend to have kind
822. 28:41of more
823. 28:43high integrity methods and again
824. 28:47they adhere to the best practices of the
825. 28:50profession
826. 28:51and i think when you can feel good about
827. 28:54that
828. 28:54you can generally generally feel better
829. 28:57about
830. 28:58what you're hearing from those reporting
831. 29:00sources because i think it's
832. 29:02pretty difficult or impossible for us as
833. 29:05individual humans to always kind of
834. 29:06track down the
835. 29:07actual like earthly point of origin
836. 29:13yeah definitely i think that's that's
837. 29:15really important and
838. 29:16give some perspective to folks who are
839. 29:19you know maybe sorting through uh
840. 29:22sorting through
841. 29:23their media literacy skills for the
842. 29:25first time
843. 29:26um wonderful we're well we've reached
844. 29:29the end of our time here together with
845. 29:31paul
846. 29:31sarah and james uh thank you all so much
847. 29:34for giving your time to address some of
848. 29:36these connections that learners made and
849. 29:38the questions that they had
850. 29:40learners thank you so much for tuning
851. 29:42into this connecting conversations chat
852. 29:44as part of the disinformation
853. 29:46misinformation and fake news teach out
854. 29:48at the university of michigan