Activity: Choose the Best Source of the Top Google Results

Here’s a simple activity you can try in your class: Have students execute a Google search that is a question. Then have the students look at the top five results, and using lateral reading pick the source that is most likely to be authoritative and the source they think is least authoritative. Have them talk through their reasoning.

For example, take this set of results to “Can magnets cure cancer?”:


We might note the first page comes from a fringe site that sells “therapeutic” magnets, and is quoting from an out-of-print magazine from the 1980s called “Magnets in Your Future” (that’s the name of the magazine, not the article). A similar critique could be made of the MagnetiCare product site.

On the other hand the two results on the bottom make a good pair. The scientific paper is cited by 25 papers, and shows in some lab conditions magnetic therapy (somewhat different than magnets) slowed progression of a specific cancer in mice. The Sloan Kettering article, however, says there is no evidence that supports the use of magnets to treat cancer. How do we weight these against one another?

It’s weird, because you would think what is reliable here would be obvious, but if you hear students talk through their thought process I guarantee you will hear things that surprise you. It may be obvious to you, for example, that a natural remedies website is less reliable than a research hospital at summarizing treatment efficacy, but to some students these may just be two different sites pitching different things — one magnets, another traditional care.

Likewise, a person with some health research literacy will know that the single study should not outweigh more general surveys of treatment efficacy, but many students will conceptualize this wrongly, not understanding that (in general) new studies serve to qualify old studies, not replace them.

You might also show the students how you think about choosing a search result yourself (assuming you’re doing it effectively!). Walk through what you do with “Does fracking cause earthquakes?”, and explain why you jump to the USGS link instead of the Forbes one, as well as why you might not see the USGS as the final word.

Need some questions to ask Google? I’ve written up over 300 for you to choose from, from ones about obvious hoaxes to ones that require a deep dive into areas where even experts conflict.

They are here.

I find pumping a wide variety of questions like this into Google helps me think about what students need when flipping through search results. Let me know if they help you do the same. Many are chosen from trending stories and a lot were generated by playing around with Google auto-suggest. I’d love at some point to rate the questions for difficulty, but right now they are just alphabetical.



Assignment: Sourcing a Quote

So this is not a photo assignment (reverse image search will get you nowhere!). But here’s a photo anyway, for aesthetic reasons:


It’s a quote from current U.S. Defense Secretary General Mattis emblazoned on a coffee mug:

“I come in peace. I didn’t bring artillery. But I’m pleading you with tears in my eyes: if you fuck with me, I’ll kill you all.”

So the question are:

  • Did he say this?
  • Who heard him say this?
  • When and where did he say it?
  • What publication was the original reporting source for this quote?

So let me say a few things about tracking down quotes (and about novice behavior when tracking down quotes). What a novice will do is this: they’ll do a web search like this for [[Mattis “i come in peace”]]:


And they’ll find a good solid publication that sources the quote:


And maybe that’s enough for daily use, but it’s not what we need here for this assignments. Quotes are some of the most bungled information on the planet. Back about ten years ago, in fact, I showed how a Washington Post story complaining about the web getting quotes wrong actually had it reversed — the web was right and the Post was wrong. (Incidentally, reviewing that post I find that it articulates pretty much what I am pushing today about web literacy — I’d forgotten how long I’d been beating this drum).

My advice for getting quotes right is the same as for everything else. Your two choices are:

  • Get as close as you can to the time and place of the quote. The original reporting, or the first reporting of the reporting. For the most part, the further you get from a quote the more it changes.
  • Alternatively, get a not just solid, but rock-solid source (such as Quote Investigator, a reputable monthly like The Atlantic, or a scholarly/major book publisher) that you know does the hard work of tracking a quote to an original source.

In case you haven’t noticed, these two options map to two of our moves:

  1. Check for previous work, and
  2. Go upstream

Anyway, go to it. Get as far upstream as you can, or to a source you consider to be rock-solid.


Assignment: Titanic Photograph

OK, I’m still on a photos kick, but this showed up in Twitter this morning:

picture of titanic

Here’s the photo alone:


And the tweet:

Was this really taken aboard the Titanic? What else can you tell us about the photo? Go to it.

As usual, don’t read the comments until after you do the assignment.

I’m going to do some text and stat based assignments next week, for folks sick of the pictures.

LazyWeb: Why Did Trust In Press Collapse in the mid-80s/early-90s?

I have a question I’d love others to answer for me.

So I was looking at longer term declines in trust in the press. And what I expected to see was a long steady fall-off from the peak trust of Watergate and if you look at some charts you see that.


But when you look closer on those charts they really elide the 1980s,

If you get granular, and look at the 1980s, the charts look like this:


That’s a dramatic drop, which coincides with Clinton’s election, and maybe with the rise of AM radio news hosts? But man — that’s really steep.


Again, you can peg the launch of Fox in here at the end (mid that elision at the end of the graph), but here there is even a significant uptick through 1985-88 (Iran-Contra?)


Here’s a different story. And again, I think about Iran-Contra for that late 1980s dive, but it just seems naive to lay it totally at the feet of one story like that. And again, AM radio is important — Limbaugh gets syndicated in 1988 and has 5 million listeners within two years. Is that enough? Or is his rapid growth more of a symptom of underlying causes here?

Let me be clear — I’m not at a loss for explanations — in fact I have many that can adequately explain this. But I’m curious if I am missing something. I’m particularly curious after reading some literature on the impact of Hollywood on press perceptions if anyone can remember popular media which may have fed this viewpoint — there’s a case to be made we don’t pay enough attention to the impact of popular media on perceptions — one can make a good argument that police have maintained good reputations despite evidence that should undermine that reputation due to a relatively sympathetic portrayal on television.

Jay Rosen looked into this a couple years back, not sure where he eventually landed (and I note that in the explanations featured AM radio, Iran-Contra, or Hollywood are not featured, so maybe I’m in the weeds here).

And of course, maybe it’s noise! It could be. But if it corresponds with something in particular that would be one route into investigation.

UPDATE: A Gallup study from the period supports the idea that at least a chunk of it was due to Iran-Contra. Brendan Nyhan — more famous recently for work on cognitive science of correction — had a solidly sourced article from 2007 arguing this is part of a pattern, with coverage of major scandals eroding trust of the governing party, and trust not returning to trend.