186 | 情境有效的执行功能评估（ConVExA）——与Yana Suchy博士的对话

欢迎收听《探索神经心理学：大脑与行为的深度之旅》，本节目由INS倾情呈现。

Welcome everyone to Navigating Neuropsychology, a Voyage Into the Depths of the Brain and Behavior, brought to you by INS.

我是瑞安·范帕顿。

I'm Ryan Van Patten.

我是约翰·博洛尼，我们都是认证的临床神经心理学家。

And I'm John Bolone, and we are board certified neuropsychologists.

今天，我们邀请到亚娜·苏奇博士，和她聊聊执行功能在日常生活中的应用。

Today, we give you our discussion with doctor Yana Suchy about executive functioning and everyday living.

亚娜是犹他大学的心理学教授，同时也是《临床神经心理学家》期刊的主编。

Yana is a professor of psychology at the University of Utah and the editor in chief of the journal, The Clinical Neuropsychologist or TCN.

她通过‘情境化执行功能评估’（ConVExA）模型的研究，已成为该领域公认的顶尖研究者、教育家和执行功能应用于日常生活的权威专家。

Through her work in the Contextually Valid Executive Assessment, or ConVExA, model, she has built a reputation as a leading research neuropsychologist, educator, and expert in executive functions as applied to everyday life.

我们觉得这一期内容和之前聊过的几个相关话题特别契合。

We think that this episode dovetails nicely with a few prior conversations we've had on related topics.

如果你感兴趣，不妨回听第118期，我们和莫琳·施密特或埃奇科姆博士谈过生态效度与数字技术，还有第138期，我们和阿黛尔·戴蒙德博士探讨过儿童的执行功能。

If you're interested, check out episode number one eighteen on ecological validity and digital technology with doctor Maureen Schmidt or Edgecombe, and episode number one thirty eight on executive functions in children with doctor Adele Diamond.

好了，接下来我们非常高兴为大家带来与雅娜·苏奇博士的对话。

And with that, we're thrilled to give you our conversation with doctor Yana Suchy.

雅娜·苏奇，真的很高兴你能来。

Yana Suchy, we are thrilled to have you.

谢谢你抽出时间参加。

Thanks for being here.

我也特别开心能来。

Very thrilled to be here.

很高兴有这个机会交流。

Glad to have this opportunity.

今天咱们主要聊两个概念：执行功能，也就是EFs，还有生态效度，不过后面我们会引入一种对生态效度的新理解方式。

Today, we're talking primarily about two constructs, executive functions or EFs, and ecological validity, although we'll talk about a new way to conceptualize the latter as we go on.

先从头说起吧，咱们怎么定义执行功能？

To start, let's clarify how we're thinking about executive functioning.

通常来说，执行功能是一组高级认知能力，帮助人们有目的地达成目标。

It's typically defined as a family of higher order cognitive abilities that facilitate goal directed behavior.

执行功能这个大范畴下包含了很多具体的概念。

There are a lot of specific concepts that fit within the broader umbrella of EFs.

研究者们提出了各种模型，我刚读研时学到的是艾黛尔·戴蒙德的理论，强调抑制控制、工作记忆和认知灵活性。但其实这个领域术语太多了——计划、组织、决策、任务切换，等等。另外，我也特别欣赏你写的那本牛津大学出版社的书，Yana，内容太棒了。

Researchers put forth various frameworks, one that I learned early in grad school was Adele Diamond's model focusing on inhibitory control, working memory and cognitive flexibility, but we know there are so many terms in this space, planning, organization, decision making, set shifting, also really appreciated your book, your Oxford book and the topic, Yana.

你能不能给我们的听众简单说说，你对执行功能的整体看法？

What would you like to say to our listeners broadly about executive functioning?

是啊，就像你说的，这话题太大了，你打算给多少时间啊？

Yeah, Like you said, it's a huge topic, and how much time do you have?

不过我会尽量简明扼要。

But I'll try to be concise.

就像你提到的，大家对执行功能的理解五花八门，尤其是学生，其实任何人聊起来，往往就是一串术语堆在一起，彼此之间却没什么逻辑关联。

Like you said, people conceptualize executive functioning in all sorts of ways, and very often what I see when students in particular, but really anybody talk about it, it sort of ends up being a laundry list of terms that aren't necessarily linked in a conceptual fashion.

比如有人会说，执行功能嘛，就是推理、解决问题、抑制冲动，然后就没下文了，一个接一个地列下去。

People will say, you know, executive functioning, it's things like reasoning and problem solving and inhibition and it just sort of goes on from there.

还有一种常见做法，干脆避而不谈它到底是什么，直接说：哦，那就是额叶的功能，对吧？

And the other way I guess people often conceptualize it is just to like get away from what it means and just say, Well, you know frontal lobe functions, right?

但当你想到其他认知过程时，我们并不会这么做。

And so when you think about other cognitive processes, we don't do that.

我们不会说，记忆嘛，就是跟时间有关的东西，也不会说，记忆就是记得去超市买东西这类事。

We don't go, you know, memory, it's like real temporal stuff, or we don't say things like, well, you know, memory, it's like remembering when you go to the store.

我们根本不需要特别解释这些。

We don't have a need to explain that.

所以我觉得执行功能这么难懂，是因为我们根本没说清楚它到底帮我们做了什么，对吧？

And so I think the reason executive functioning is so difficult is that we don't have a clear conceptualization of what it does for us, right?

记忆的进化是为了帮我们记住事情，那执行功能进化出来又是为了什么？

Memory evolved to help us remember, so what does executive functioning evolved for?

所以我喜欢把它看作是运动系统的延伸。

And so I like sort of to think about it as this sort of outgrowth of the motor system.

我们从最原始的生物开始，只有刺激-反应，发生什么事，运动系统就直接回应，根本没得选。

You know, we start out with primitive species with stimulus response, and so something happens and the motor system responds to it and there is no choice in what you're going to do about it.

比如飞蛾，一次又一次地扑向火焰，烧了翅膀也不长记性，对吧？

So if you're a moth, you're going to fly to the flame time and time again and burn your wings and don't learn from it, right?

没有记忆，也没有执行功能。

No memory, no executive functioning.

随着物种变得越来越复杂，它们就开始有了选择。

And so as species get more complex, they begin to have choices.

拥有这些选择，在进化上是有优势的。

And so there is evolutionary advantage to having those choices.

所以我觉得，执行功能其实就是让我们能做选择的那项能力。

And so I think we have to think about executive functioning as something that gives us that advantage of choices.

它真正发挥重要作用的地方，是在社会行为上——有意思的是，社会性物种的执行功能往往更发达。我常举的例子是锤头鲨，它们算是比较原始的物种，但很社会性；而其他鲨鱼比如大白鲨就不那么社会，它们的前额叶却小得多。

And where it really begins to play an important role is when we think of social behavior, and it's interesting that social species have more evolved executive functioning, and it's like a great example I like to use is that hammerhead sharks who are pretty primitive species but they're social, whereas other types of sharks are not, and they have much bigger frontal lobes than say the great whites, right?

生活在群体里，你就得能做选择。

And so you have to be able to make choices when you live in a social group.

你得能压制住哪怕是很强烈的冲动。

You have to be able to inhibit very even strong urges, right?

比如大猩猩，你得服从银背首领，不能见谁就往上扑。

If you're a gorilla, you have to defer to the silverback and not just go mate left and right with anybody.

所以，能掌控自己的行为、抑制冲动和欲望，这能力真的特别重要。

So it's like pretty important ability to make choices about your behavior and inhibit impulses and urges and that kind of stuff.

随着物种越来越复杂，尤其是变成人类之后，我们这些行为选择——全靠执行功能支撑——也变得越来越复杂。

And so as you begin to be more and more complex species, right, you become a human being, now these behavioral choices that we have and that executive functioning gives us become more and more complex.

我们不再只是像实验室里的老鼠那样，简单决定往左还是往右走，而是开始把选择推得更远、更提前，还涉及越来越多抽象层面的东西。

We no longer just decide whether to go left or right, right, like a rat in a maze, say, but we begin to make those choices further and further forward and also in terms of more and more sort of abstract types of things.

你看，随着前额叶越来越往前发展，它帮我们做的决策也越来越面向未来、跨越时间，也越来越抽象。

And so you really see how even as the frontal lobes grow sort of forward, the things that the frontal lobes can do for us in terms of decisions become more and more forward, like distance and time oriented and more and more abstract.

比如，选好吃的还是不好吃的，这种决定靠的是前额叶靠后一点的区域；但像‘我到底要不要读完研究生’，或者‘钱’这种概念性的选择，就更靠前、更抽象了。

So just when you think about choices between something that tastes good and doesn't taste good, that happens sort of more posteriorly in the frontal lobes, whereas choices between do I want to finish graduate school, that becomes more abstract or concepts even like money and stuff.

说到底，在我看来，这原本只是些简单的运动反应，后来却演变成了一套能让我们做出越来越复杂决策的高级机制。

And so it's really, in my mind, what started out as motor responses and suddenly became this complex construct that allows us to make choices that are more and more complex.

归根结底，就是怎么从A点走到B点的问题。

And so ultimately, it's about making it from point A to point B.

我一向这么理解它。

That's how I like to talk about it.

比如你早上起床时有计划，到晚上真能达成那个目标吗？

Like you get up in the morning, are you at night going to end up where you planned to end up?

那为了实现它，你到底该做点啥？

And so what do you need to do, right, in order to accomplish that?

你肯定得做那些大家一提到执行功能就想到的事儿——推理、解决问题、做计划，但这些全都是脑子里的事。我总觉得，很多人谈执行功能，光想着‘我规划了’‘我解决了问题’‘我把东西整理好了’，可你人还瘫在沙发上，脑子里想得挺美，事儿一样没干成。

You need to for sure do the things that people often think of when they think of executive functioning, so reason, problem solve, plan, all those things though are just in your head, and it strikes me that very often when people talk about executive functioning, they just think about the things that go on in your head, like, oh yeah, I planned, I solved a problem, I organized what I need to organize, And forget that right now you're still sitting on your couch planning in your head and you haven't accomplished anything.

你压根儿没行动啊，对吧？

You haven't executed, right?

这叫执行功能，不是光想就行。

It's called executive functioning.

你得先想出个计划，然后得真启动起来，付诸行动。

So you first need to come up with this plan, and then you need to initiate initiate behavior.

行动。

Behavior.

一旦你开始行动了，还得持续下去，而且得管住自己，别干那些跟目标背道而驰的蠢事。

Once Once you you initiate initiate behavior, you then need to maintain that behavior appropriately, and that means inhibit inappropriate actions that are not consistent with your goals.

所以你有这么一个认知层面，执行功能包括行为的启动和维持，还有对行为的抑制。

And so you have, you know, this cognitive piece, executive cognitive functions, have initiation and maintenance of behavior, inhibition of behavior.

我们有时候忘了，目标不只是到达B点，还得继续往C、D、E、F……一路走下去。

The thing that we sometimes forget about is that it's never just about making it to point B, It's also making it to point C and D and E and F and so on and so forth.

所以要想在一天结束时完成这么多目标，你还得合理地把各个任务的子目标穿插安排好，而要做到这点，就得反思自己的能力——哪些是现实的，哪些能办到，哪些办不到。

And so in order to accomplish all those multiple goals by the end of the day, you also need to sort of organize or interleave appropriately sub goals for one task, sub goals for another task, and in order to do that appropriately, you need to reflect on what your capacities are, what's realistic, what you can or cannot accomplish.

这就涉及前瞻记忆和自我反思的能力，比如：我是不是得给自己设个提醒？还是说记住X、Y、Z这些事本来就在我的能力范围内？

And so this ability to engage prospective memory and self reflection insight on, for example, well, do I need to write a reminder for myself or is that within my capacity to remember X, Y, and Z?

这其实是在我们之前说的那些——解决问题、启动行为、维持行为、抑制行为——之上，又叠加了一层更复杂的机制。

So it's this additional more complex layer that's sort of superimposed on the things we talked about initially, you know, problem solving and initiation and maintenance and inhibition of behavior.

你还需要具备这种更高阶的能力。

You need to also have this higher level yet.

这就像执行功能的执行功能。

It's like the executive functions of the executive functions.

这是一种更高层次的能力。

It's like a higher level ability.

所以这就是执行功能帮我们做的事，对吧？

And so that's what executive function does for us, right?

如果我们只把执行功能拆成这些小块——抑制、解决问题，那我们就忽略了真正该评估和考虑的整体图景，还有这些能力的缺失会怎样影响日常生活中的行为。

And so if we just break it down into these little pieces, it's inhibition, it's problem solving, I think we're missing the big picture of what we need to be assessing and considering, and also how that translates, right, breakdown in any of those processes, how that translates to everyday behavior.

我喜欢你提到的这种层级结构，这对执行功能来说确实特别关键。

I like how you talk about the hierarchy that's certainly super relevant to executive functions.

我想到一本以前读过的书，作者好像是胡安quin·富斯特，讲的是前额叶皮层。我在研究生时看过一张图，展示了神经组织的层级：最底层是脊髓与肌肉的突触连接，产生动作；往上走，是单模态联合皮层、多模态联合皮层，越来越高级，整合的信息也越来越多。

I'm reminded of a book, Joaquin Fuster, I think it is, it's on the prefrontal cortex I read in grad school where there's a nice diagram showing on the one hand you have the hierarchy of neural organization, At the lowest level, we have muscle activation, you know, synapsing from the spinal cord on a muscle and we make a movement and you move up that hierarchy to unimodal association, polymodal association, cortices in the brain, higher level integrating more information.

同样地，行为也有个层级：从最简单的运动激活，一路往上，到最复杂的执行功能。

In parallel, we have, you know, simple behaviors and then moving up from what we might consider, you know, very straightforward motor activation all the way up to the most complex executive functioning.

但正如你所说，它本质上是行动导向的。

But it is inherently action oriented, as you mentioned.

是的。

Yeah.

当然。

Absolutely.

我们常常忽略了自己的行为有多复杂，其他动物也能做些特别厉害的事，但它们没法像我们这样灵活应对，也不能像我们这样随机应变。

And we sometimes take for granted how complex the behaviors that we are engaged in are and how other animals can do pretty amazing things, but they can't do them flexibly in the same way we can and can't adapt in the same way we can.

就像你提到的，社会压力确实推动了语言的发展，也让我们不得不应对这种复杂性。

Like you mentioned, the social pressures that we have really drove the need for language and this complexity that we deal with.

所以我们每天都在理所当然地使用它。

So we take this for granted just every day.

对大多数人来说，这事儿来得很容易，但从神经解剖学角度看，其实特别耗脑力。

It comes easy to most of us, but it is really neuroanatomically taxing.

接下来我们想再多聊聊这个话题。

And so we wanna talk more about that throughout the rest of the conversation.

执行功能的测试，跟其他认知领域的测试一样，也有类似的发展历程：早期这些测试主要是为了定位大脑损伤，但随着神经影像技术的进步，这种需求就没那么大了。

And and the test of executive functioning, have a similar historical trajectory to tests in other cognitive domains in that in the early days, our tests were used primarily to localize lesions in the brain, but with the advancement of neuroimaging techniques, this is not as needed.

后来整个领域逐渐转向用这些工具更全面地描述认知功能，预测人们在日常生活中的表现。

And we transitioned, the field transitioned toward using our tools to spend more time characterizing cognitive functioning broadly, predicting everyday functioning.

当然，现在还是有些情况会用测试来检测脑部病变，比如跟阿尔茨海默病的关联，或者术中脑功能定位，就是几个例子。

There are obviously still scenarios where it is useful to use tests to detect brain pathology or lesions like correlations with Alzheimer's disease or intraoperative brain mapping procedures, is a couple examples.

但我们的领域似乎正越来越转向预测日常生活中的表现。

But our field is really, it seems to be moving toward the prediction of daily activities.

我想知道你是否认同，以及你怎么看待我们这些测试所承担的不同作用。

I'm curious if you agree and how you think about these different purposes that our tests serve.

是的，当然。

Yeah, for sure.

我觉得，我们在报告里已经不会说‘这个测试的表现提示右颞叶有病灶’之类的话了。

I think, you know, in terms of what we say in our reports, we don't say, Oh, this, you know, performance on this test suggests that there is a lesion in the right temporal whatever.

现在没人这么说了。

We don't say those things anymore.

我常跟我学生讲：如果你报告里满篇都是大脑结构，那说明你根本不是临床神经心理师，只是在装样子。

And in fact, I always say this to my students, like if you talk a lot about the brain in your report, it suggests you're not a neuropsychologist, you're just sort of faking it.

这根本不是重点，对吧？

Well, it's not about that, right?

现在找我们做评估的人，根本不在乎这些。

That's not what the referral sources want anymore.

以前是，但现在不这么做了。

They used to, but they don't anymore.

不过，你还是得像以前一样，理解大脑和行为之间的关系，能把测试表现和诊断相关的脑网络联系起来，对吧？

That said, you still need to have the same understanding of brain behavior relationships like you always did and be able to correlate performance on a test to neural networks that are relevant for diagnostic decision making, right?

虽然我们在报告里不提这些，但依然得从脑网络或认知过程受影响的角度，去分析某项测试的表现，或者多项测试的整体模式。

So while we don't talk about it in our reports, for example, we still definitely want to be able to think of performance on a test or a profile across the tests in terms of the networks that are affected in the brain or what processes are affected.

所以在做诊断决策时，我们确实一直都在这么做。

And so for diagnostic decision making, we for sure continue to do that.

但我还想说，你同样得时刻考虑这些能力缺陷或局限对日常生活的影响，这一点对诊断同样至关重要。

But I would also argue that you similarly need to always think about the ramifications of any of those, say weaknesses you may have or limitations for daily functioning, and that too is actually critical for diagnostic decision making.

比如诊断痴呆，就必须证明患者已经无法独立完成日常生活事务。

So for example, for the diagnosis of dementia, have to demonstrate lack of ability to independently function in daily life.

再比如智力障碍，诊断不能光看分数——很多人以为神经心理学家就只会看数字，其实那只是我们工作的一小部分。

Or for intellectual disability, you have to really base that diagnosis not on the numbers, right, where just sometimes people think of neuropsychologists, that it's all about the numbers, and I feel like that's just really a sliver of what we do.

还有像注意力缺陷多动障碍这些情况。

So there are, you know, ADHD.

ADHD其实根本不需要那些分数，关键得靠面谈，看人在日常生活里到底怎么样。

ADHD, you actually don't even need the numbers, you really have to have the interview, and again, how people function in daily life.

我觉得这两方面都特别重要，不管是诊断还是我们提的日常功能建议，而且这方面的需求确实在不断增长。

So I think both are critically important, both for diagnosis, but then also for those recommendations we make in terms of daily functioning, and for sure that is a growing area.

但我不会说我们完全抛弃了脑行为关系这套东西，还是得搞清楚哪些脑网络跟哪些病有关，这些基础不能丢。

But I wouldn't say that we are switching completely to that away from those brain behavior relationships and understanding what networks are related to what diseases and that type of stuff.

我注意到，这些年写神经心理报告时，我越来越把重点放在‘印象’和‘建议’部分，尤其是关于患者真实生活中的表现。

I've noticed over time in my neuropsych reports, I've moved more and more toward spending time in the impression section and certainly in the recommendations on real world everyday functioning.

有时候还是会用认知轮廓来区分阿尔茨海默病、血管性痴呆或者路易体痴呆，这些临床特征在某些情况下还是有点用的。

There's there's still a little bit at times of using the cognitive profile to do a differential diagnosis between Alzheimer's disease and supravascular disease or dementia with Lewy bodies, the clinical profiles that can have a little bit of utility in some situations.

这其实还能扯出一大篇，比如区分阿尔茨海默和血管性痴呆到底有没有必要？

It's a whole another conversation we could get into, you know, is it useful to differentiate AD versus vascular?

我们的测试在这方面通常都做得不够好。

Our tests often don't do that super well.

就像你说的，我们真正擅长的、也更有价值的，是预测预后——不管这个人到底有哪些合并病理，这些测试能告诉我们，这到底意味着什么。

As you said, what we can do well, better, I think, and is so much more useful as prognosis and what does this mean regardless of which co pathologies are involved in this particular person, which we can't fully disentangle with our tests.

这对他工作、上学，或者生活中最重要的事意味着什么？

What does it mean for work and school and whatever activity is most important to that person in their life?

这正好引出你的模型——情境化执行功能评估，也就是ConVExA模型，它把执行功能和各种特质因素、情境影响联系起来，比如睡眠、疼痛、情绪压抑，这些概念都很有意思。

So that that's a good segue into your model, the Contextually Valid Executive Assessment or ConVExA model, including in the interaction between executive functioning and various trait based factors and situational influences, sleep, pain, expressive suppression, interesting construct.

能给我们讲讲这个模型吗？它是怎么把执行功能和日常生活联系起来的？

So tell us about the model and what it how it sort of links executive functioning and everyday living.

当然可以。

Oh, sure.

我觉得我们都同意，执行功能对日常生活至关重要。

So, you know, I think we would all agree that executive functioning is critical for daily functioning.

那我们就从这个共识开始吧。

So I think we can just start with that assumption.

这个模型的核心观点是，执行功能和日常生活表现都是会波动的。

The model is based on the notion that both executive functioning and functioning in daily life fluctuates.

当你在诊室或实验室评估执行功能时，你看到的只是这个人‘特质层面’的一个 snapshot，但实际数据会有波动，这不光是因为测试不够精准，更因为这个人当下的状态本身就在变——他的执行能力在那一刻是起伏不定的。

So when you assess executive functioning in your office or in your laboratory, you're getting a bit of what the trait level is, but there is going to be a standard deviation around it, right, that has to do not just with the lack of precision of our tests, but also with where that person is at the moment, because they themselves are fluctuating in terms of their executive abilities in the moment.

所以，正如我所说，执行功能跟日常生活息息相关，那日常表现也会有波动，这很自然。

And so since, like I said, executive functioning relates to daily functioning, it makes sense that daily functioning also fluctuates.

我觉得这点大家肯定都有体会。

And I think we can all relate to that.

我们都有状态好的日子，也有状态差的日子。

We have good days and bad days.

有时候啥都不顺，事儿事儿的；但也有时候特别顺，思路清晰，干啥都顺手。

We have days where we just, everything goes wrong, right, and other days when we're really clicking.

所以，如果我们真想在临床中准确预测患者的日常表现，就得把这种波动性考虑进去。

And so I think we need to incorporate this notion of fluctuation into our prediction of daily functioning, if that's what we want to do clinically and with our patients.

那哪些因素会导致这种波动呢？比如你提到的疼痛、睡眠不足，还有‘情绪压抑’这个概念——我们实验室最近特别关注这个，它其实是一种不太健康的情绪调节方式，跟心理治疗里教的那些方法不一样，简单说就是憋着不表现情绪，对吧？

And so what are some of the factors that lead to fluctuations, like you said, things like pain, for example, lack of sleep, the construct of expressive suppression is actually an interesting one that we've been looking at quite a bit in our lab, and it has to do with, it's a type of emotion regulation that's not as healthy as the types of emotion regulation that we teach people in psychotherapy, it has to do with basically sort of hiding your emotions from others, right?

就是压住脸上的表情，或者克制住情绪外露的行为。

So suppressing your facial expression or behavioral expressions of your emotions.

结果发现，这特别耗能量。很多研究，甚至不光是神经心理学的，都表明：如果你让人憋着情绪一段时间，等结束后，他们的执行功能表现会明显变差。

And it turns out it's quite taxing, and quite a bit of research, even outside of neuropsychology, shows that if you tax people's expressive suppression, if you have them suppressed for a period of time, they will not just end the time but subsequently perform more poorly on measures of executive functioning.

所以重点不光是 multitasking，同时做两件事，两样都得靠执行功能，比如在测试里还要憋着不露表情，这其实会大量消耗你的心理资源，结果就是你后面表现还是差。

So the notion is it's not just about multitasking, doing two things at once, right, both having to be executively involved, say in a test and suppressing your facial expressions, but it actually taxes your resources and subsequently you still perform more poorly.

我们研究得挺多的，真挺惊人的——哪怕只是让人看几张恶心的照片，要求他们五分钟内别露出厌恶的表情，就足以耗尽他们的执行功能，五分钟之后表现就明显下滑了。

And so, you know, we've been looking at this quite a bit and it's pretty striking actually that even something as silly as showing people like disgusting pictures and asking them not to make a disgusted face for five minutes is enough to actually deplete their executive functioning and the performance is affected five minutes later.

所以这就是我们正在研究的一些概念，我们也知道，比如疼痛也是类似：人要是正忍受疼痛、还得努力应对，之后一段时间内，他们的执行功能就会被耗光。

So that's just some of the concepts that we're looking at and we do know that, for example, similar research has shown that with pain, if people are experiencing pain and having to cope with the pain, then subsequently for a period of time they are depleted in terms of their executive functioning.

所以这个模型的核心是：有各种因素会在一段时间内影响并耗尽执行功能，而这些因素正是通过消耗执行功能，间接影响了日常表现。

And so the model basically says that there are these various factors that can impact and deplete for a period of time executive functioning, and so these factors impact daily functioning, but this impact is mediated by executive functioning that's being depleted.

这个模型还有另一个层面，不光是‘中介’，还有‘调节’作用。

There is another aspect of the model that looks at not just mediation but also moderation.

也就是说，执行功能本身也会和这些因素相互作用，共同影响日常功能。

So executive functioning also interacts with some of these factors to impact daily functioning.

举个例子，如果一个人日常饮食缺蛋白，生活特别忙，每天都不一样，得天天动脑筋想办法怎么撑过这一天，这会直接影响到他们的日常生活能力，对吧？

So as an example, if a person has lack of protein in daily life, very busy, every day is different, from one day to the next you have to sort of problem solve how you get through the day, that impacts your instrumental activities of daily living, right?

这其实很好理解。

And that makes a lot of sense.

你的时间更少，出错的空间也更大。

You just have less time and less, you know, there's more room for error.

但结果发现，这种关系会受到执行功能的调节——执行功能特别强的人，生活再乱也能把事都办妥；但哪怕你只是普通人，缺乏规律的生活也开始显出代价了。

But turns out this relationship is moderated by executive functioning, so people with very high executive functioning can have as messy a life as they want and they'll still get everything done, but even if you're just average, you already begin to see a toll to that lack of routine.

所以，执行功能特别强的话，这时候能保护你。

And so very high executive functioning protects you in this case.

再比如，我没在研究里验证过，但我猜这可能是个曲线关系：当执行功能严重下降时，有没有规律其实都无所谓了，反正你啥也干不成，对吧？

Then if you were to go, there's probably, I haven't tested this and looked at this in research, but I suspect there is a curvilinear relationship such that when the executive functioning becomes quite poor, then the lack of routine doesn't matter anymore because you can't do anything regardless, right?

所以我们得始终结合具体人群来理解这些关系。

So we have to always think about any of these relationships in the context of the population you're looking at.

举几个例子，比如当我们让人刻意压抑情绪表达时，他们的任务表现就会受影响，而这种影响是通过执行功能被消耗来实现的，对吧？

So just a couple of examples, for example, you know, when we expose people to expressive suppression, then that will impact their performance on ideal tasks, but that relationship is mediated by executive functioning, which is also impacted, right?

这是其中一种模型。

That's one model that you look at.

另外，像我之前说的蛋白质那个例子，我们现在正把同住情况——是独居还是和配偶一起住——当作一个环境因素来研究。

And the other thing is, like I said, the business like a protein, we're just looking right now at cohabitation actually as a sort of a contextual factor, whether you live alone or with a spouse.

再说一遍，这些是老年人，如果他们的执行功能高于平均水平，有没有伴侣对他们的日常活动能力都没影响。

And again, turns out that people who, these are older adults, so if they're sort of above average on their executive functioning, their performance of IDLs is unaffected by having a partner.

但哪怕你只是平均水平——这些也是老年人，换算成量表分数，大概相当于30多岁的人得8分左右，所以我们得放在这个背景下看。

But even if you're just average, so these are older adults, so that would translate to scale scores of like eight in people in their 30s, so again, we have to put it in that context.

但如果你都70多岁了，执行功能只是普通水平，和配偶一起住能大大提升你在家独立完成日常事务的能力。

But if you're like average and you're in your 70s, living with a spouse hugely improves your ability to perform IADLs independently at home.

这种调节和中介效应，正是这个模型关注的重点。它的核心观点是：我们在诊室里测出一个分数，并不能直接预测一个人的日常生活表现，因为现实中有太多波动和其他因素在起作用。

So those types of moderations and mediations is what this model looks at, and the notion is that having a number pop out of our tests in our office doesn't tell us how to predict daily functioning because there will be these fluctuations and these other factors that will impact that.

是的。

Yeah.

我的执行功能肯定挺差的，因为我根本没法不受老婆的邋遢影响，日常生活全被她带偏了。

My executive functioning must be pretty poor because I cannot help but, have my wife's disorganization affect my my everyday functioning, so.

约翰，我本来想说，这个模型或许能解释你为啥生活这么有条理、计划性强——其实是在弥补你执行功能偏低的问题。

John, I was gonna say the, the model might explain why you're so organized and planful in your life to compensate for your low executive functioning.

我以前一直不明白为啥你这样，现在ConVExA模型总算说清楚了。

I never knew why and now the ConVExA model tells us.

我呢，还得天天憋着不冲你发火，因为你老让我干些无聊的破事。

Well, I'm also I'm also dealing with constantly suppressing my expressing rage at you for asking me to do some other dumb task.

这事儿我也得天天跟它较劲。

I have that to battle with as well.

但雅娜，我想问问，个体内部的波动性——毕竟这些能力每天甚至一天之内都会变来变去——这在执行功能里是不是特别重要？

But Yana, I was wondering how intra individual variability, because these skills vary day by day or even within a day in each of us, how might that be an important aspect of executive functioning?

是啊，这问题问得好。目前研究还没搞清楚，测试中的波动和日与日之间的波动是不是一回事，但我猜它们可能不一样，不过我觉得背后有个共同的机制在起作用。

Yeah, that's a great question and of course the research still remains to tell us whether fluctuation within a test is the same as fluctuation from day to day, and I suspect they may be different, but I think there is a mechanism that's linking both.

我们确实知道，比如执行功能能帮我们抵御一些外部环境因素的影响，就像我刚才说的。

So we do know, for example, that executive functioning protects us from some of these executive factor, contextual factors, as I just was saying.

所以很自然地，你的执行功能越强，每天的表现就越稳定，对吧？

And so it would make sense that the higher your executive functioning, the less you fluctuate from day to day, right?

从这个角度看，我觉得这两者之间肯定有很明确的关联。

So from that standpoint, I think there is probably a pretty clear association.

至于测试过程中的波动，比如你做反应时任务时，突然有那么几下没跟上节奏，这通常也被解释为执行功能不足，也就是注意力控制偶尔出现小漏洞。

Now, fluctuations say within a test, so if you do a reaction time task and you have little blips where you suddenly sort of don't go as fast as you could have been going, that's been typically interpreted also as a weakness in executive functioning, as, you know, little sort of lapses in your attentional control.

这很有道理，研究也支持这一点，从这个角度看，它们确实有关联，但我觉得它们反映的是不同层面的问题。

And that makes a lot of sense, research supports that, and so in that sense they are related, but I think they capture things at a different level.

所以，任务内部的个体差异，可能更触及注意力控制的细微之处，而另一种则更多反映你抵御环境干扰的整体能力。

So the inter individual variability within a task probably gets at these subtleties of attentional control versus the other one gets more at these grosser level associations with how well you protect yourself from the impact of the contextual factors.

我不确定该不该说，不过我想提一点：我们接受训练时都学过，应该忽略一次糟糕的分数，对吧？

Not sure if I should say, I guess one thing I would say about that too is that I think we all learned in our training that we should ignore one bad score, right?

毕竟，谁都会犯错。

So like anybody can make a mistake.

这个想法是，人会紧张，会压力大，会出错，但我想问的是：如果在受控环境下，你明知道没那么紧张，还是出错了，那在真实生活中呢？

This notion is like anybody can make a mistake, and there's the pressure of being tested and you're anxious, and to me it's like, well, if you make a mistake in this protected environment, but anxiety notwithstanding, then what do you think happens in daily life?

所以我觉得，我们真得认真看待测试中的这些小失误。

And so I think we do need to interpret those lapses on our tests.

我们不能因为一个人一次没考好，就断定他的执行功能很低，对吧？

We can't interpret a lapse as, you know, if somebody flunks a test, we can't say their executive functioning is at this low level, right?

不能一棒子打死说‘他有障碍’。

Like, oh, they're impaired.

不，他们没受损，但这告诉我们，他们也会出现疏忽。

No, they're not impaired, but what it does tell us is that they can have lapses.

他们也会有走神的时候，说白了就是考砸了。

They can have moments when they're not online, right, and they flunk, so to speak.

如果他们能在测试中出错，那在日常生活中也可能出错，比如忘事儿、丢东西——甚至有悲剧发生，有人把婴儿忘在车里，对吧？

If they could flunk a test, they can flunk a daily task and, you know, they leave, forget, I mean, there are tragic examples when people leave an infant in a car, right?

也就是说，一瞬间的疏忽，可能带来可怕的后果。

I mean, so it's a momentary lapse that can have horrible consequences.

所以，虽然我们不能把分数高低直接等同于执行功能的水平，但这些错误确实值得我们重视。

And so while we can't interpret the level of the score as being the level of capacity for executive functioning, I think we do need to consider those errors.

其实有意思的是，我们做过一些研究，专门看执行功能测试中的错误，这些错误本质上就是疏忽，是表现的波动，比如在像A试验这种简单任务里也会出错。

And actually what's interesting, we've done some work with looking at, for example, errors on measures of executive functioning, and again, you can think of errors as lapses, as fluctuations in a way, and lapses on like the simple tasks like trials A.

谁会在A试验里出错啊？

Like who would make an error on trials A?

可就是有人会。

People do.

结果发现，你在A试验中出错，并不是因为你不会从1数到10，对吧？

And turns out that if you make an error on trials A, it's not an error in your ability to go from one to 10, right?

不是你不会按顺序来，而是执行功能出了问题，是注意力脱节了。

It's not an inability to sequence, but it's an executive lapse, it's an attentional lapse.

所以，这些简单任务里的错误，跟那些复杂任务里的错误一样，都属于执行功能的问题。

And so errors on those simple tasks are just as executive as errors on the executive tasks.

所以，这是一种波动，对吧？

And so that's one type of fluctuation, right?

错误就是一种波动，尤其是在简单任务上，我们临床上真的得认真看待。

Errors are a type of fluctuation, especially on easy tasks, that we do need to interpret clinically.

我觉得还得做更多研究，才能准确解读这些现象，但我觉得不能直接说‘谁都会犯错’就全忽略掉。

I think more research is needed as to how exactly to interpret it, but I don't think we should completely ignore it and say anybody can make a mistake.

我们既不能因为几个低分就过度解读、过度病理化病人，也不能完全无视这些低分，尤其是结合病人的具体情况——他们的工作、日常生活中的失误。错误的类型本身可能就揭示了，到底是什么问题让病人最初走进了神经心理诊所。

Not over interpreting or over pathologizing a patient based on a few low scores, but also not totally discounting those few low scores, particularly in light of our patient's characteristics, their work, you know, mistakes they're making in everyday life, and depending on what the error is, it could be illuminating with respect to what's going wrong for the patient in their life that led them into the neuropsychologist's office in the first place.

对。

Yeah.

说得对。

That's a good point.

那我们接着说，来聊聊生态效度这个很有意思的概念。

So now moving forward, let's bring in ecological validity, this very interesting construct.

首先，生态效度主要有两个核心部分，听起来有点专业，但其实挺重要的。

First, there's these two primary components of ecological validity, a bit esoteric but also important.

一个是表面真实性，也就是测试看起来多像现实生活中的场景，有点像表面效度。

There's verisimilitude, which is the extent to which a test appears on its face to resemble real world scenarios, so it's similar to face validity.

另一个是实质真实性，指的是测试实际预测日常生活表现的能力，更接近预测效度。

And then there is veridicality, which means the degree to which the test actually predicts everyday functioning closer to predictive criterion validity.

明白了这两个概念后，我想听听你对当代神经心理学中生态效度这个概念有什么看法，你觉得应该用什么来替代它？

With these terms in mind, I'm curious to hear you talk about problems with ecological validity as a construct in contemporary neuropsychology and what you recommend instead.

是的。

Yeah.

我觉得‘生态效度’这个词最早出现，是因为大家发现，我们现有的执行功能测验虽然能测出抑制、解决问题、行为维持这些层面，但根本碰不到我前面说的那个更高阶的层面——也就是需要整合自我觉察、前瞻记忆和元认知监控的那个层面。我们现在的测试，通常根本测不到这个高阶能力，对吧？

You know, I think the term ecological validity first emerged when people realized that our measures of executive functioning kind of get at that inhibition, problem solving, maintenance of behavior level, but don't get at that higher level I talked about earlier where you need to incorporate insight and perspective memory and sort of monitoring at a meta level, this higher level does not typically get assessed by our tests, right?

所以如果你在我们那些比较直接的测试里表现不好，比如斯特鲁普测试或者类似的任务，这些测试主要测抑制能力，但你到了更高层次的元认知层面，比如整合洞察力、自我监控这些能力时，可能还是会出问题。

So if you perform poorly on our typical measures that get more straightforwardly, let's say inhibition like the Stroop test or some variants of that, then once you have problems successful at that meta level, that higher level.

但反过来，就算你在传统测试里表现很好，也不代表你在这些更高层次的认知过程上就没问题。

But if you perform well on our traditional tests, you could still have a problem with these higher level processes.

而我们的测试根本测不到这些，所以当初人们开始质疑执行功能测试缺乏生态效度，就是从这儿开始的。

And our measures don't assess that, and so that was where originally these complaints about measures of executive functioning not having ecological validity began.

‘生态效度’这个词听起来挺时髦的，直觉上也特别有道理，所以很快就流行开了。

So the term ecological validity is sort of a sexy term and it has sort of real, like intuitive nature to it, and so it caught on.

其实这个词最早是实验心理学家用的，他们讨论的是实验中发现的那些关系，能不能真正延伸到日常生活中去。

It actually, the term originates from experimental psychologists talking about whether experimental relationships, you know, relationships on their various experiments, whether they translate to daily life.

后来这个‘从实验室推广到现实生活’的概念，慢慢被用到了各种认知测试上，尤其是执行功能的测试，用得最多。

And so this notion of generalization from the laboratory to daily life then sort of was beginning to be applied to cognitive tests in general and then to tests of executive functioning in particular most often.

再往后，大家对这个词的理解就变成了：测试得看起来像现实生活才行。

And so this got on and the way people began to interpret it was, well, the test should look like daily life.

结果这想法都快走火入魔了，比如有人说：别做连线测试B了，咱们改用做三明治的图片吧，或者换点别的现实场景的刺激——好像只要把测试里的抽象符号换成真实生活里的东西，就能让结果更普适了。

And that actually can get as silly as, saying, oh, instead of doing trails b, we'll have pictures of, you know, sandwich making, or I don't know, basically, like, exchanging the stimuli in our current tests for stimuli from the real world instead of abstractions somehow would make things more generalizable.

其实还有别的办法，可以弥补我们传统测试里缺失的那些方面。

There are actually other things one can do to introduce those issues that were lacking in our traditional tests.

比如，测量前瞻记忆的测试就是一个方向。

So prospective measures of prospective memory is one way to go about it.

另一种方法是减少结构化的限制，要求被试同时运用前瞻记忆和这种高阶的组织能力。

Another way is to really take away some of the structure and require that prospective memory along with this sort of meta level organization be applied.

BADS量表里的‘六元素测试’就是个很好的例子。

And so for example, the Six Elements test from the BADS battery is a great example of that.

这些测试其实从80年代末、90年代初，人们开始质疑生态效度的时候，就马上被开发出来了。

And so those tests pretty much were developed right off the bat as the complaints about ecological validity started in the late 1980s, early 1990s.

但不知怎么的，整个领域总觉得还得继续搞更多‘生态效度’高的测试，这在神经心理学界简直成了一股风潮。

However, the field somehow felt that we need to keep developing more ecologically, quote unquote, ecologically valid tests, and so this has been, I want to say, bit of a craze in neuropsychology.

搞出一个看起来跟现实生活一模一样的测试，听起来特别酷，好像这样就能完全捕捉到日常表现似的。

It seems very sexy to come up with a test that just looks exactly like the real world and somehow that test will then capture daily functioning.

但我的看法是，就像我们之前说的，日常表现本来就是波动的，执行功能也是，根本不可能靠一个测试就完美预测一个人的日常表现。

Now, my response to that is, as we discussed, daily functioning will fluctuate, executive functioning will fluctuate, so there is no one test that will somehow perfectly predict daily functioning.

大多数人觉得，如果测试看起来不像真实生活，那它也就没法预测真实生活中的表现。

And most people assume that if the tests don't look like the real world, they also cannot predict real world functioning.

这个术语被搞得一团糟，连神经心理学之外的人都在抱怨，说要么干脆别用这个词，要么至少用的时候必须明确定义。

There is so much confusion about this term that outside of neuropsychology, more broadly, there have been people complaining about the usage of the term and advocating that it should be either eliminated or at least making sure that it's always clearly defined when being used.

在我们实验室，其实我们也发了几篇论文，核心主张就是：别再用‘生态效度’这个词了。

In my lab, we have actually also, we have a couple of papers out there where we, our call to action is let's get rid of the term ecological validity.

原因很简单——我们根本不需要它，它不但帮不上忙，还老惹麻烦。

The reason being is that actually we don't need it, it does not actually help anything, it introduces confusion.

如果你想说测试有多像真实场景，那直接说‘表面效度’不就完了？

So if you want to talk about verisimilitude and what the test looks like, why not just say face validity?

如果你想说测试覆盖了哪些认知过程，那就叫‘内容效度’呗。

Or whether the test taps into some subset of processes, you could call it content validity.

其实我们早就有现成的、好用的术语了。

So we already have perfectly good terms for that.

如果你想讨论测试能不能预测未来表现，直接说‘预测效度’或者‘准则效度’不就得了？

If you want to talk about the test's ability to predict some future outcomes, you can talk about predictive validity or certain criterion, criterion validity.

我们其实早就有了非常清晰准确的术语来表达这些概念。

Again, we already have perfectly good terms that very clearly express those ideas.

但一用‘生态效度’这个词，就会把本来各自独立的几个概念搅在一起，导致大家得出错误的结论。

But when you use the term ecological validity, it begins to confound all these various pieces that, you know, are thought to contribute to that term and lead to erroneous conclusions.

很多同行都坚信，执行功能测试或者传统测试根本预测不了日常生活中的表现。

So there is a lot of belief out there among our colleagues that tests of executive functioning or traditional tests do not predict daily functioning.

可实际上，翻翻文献你会发现，大量研究得出的结论恰恰相反。

And turns out, when you look at the literature, there is a huge body of literature that says otherwise.

这不只是零星的研究，而是有系统性综述和元分析，全都指向同一个结论：我们的测试确实能预测日常生活表现。

It's not just individual studies, there are systematic reviews, meta analyses, and they all come to the same conclusion that our tests do predict daily functioning.

它们能预测得完美吗？

Do they predict perfectly?

不能，根本不可能。前面说过，测试本身总有误差，再加上执行功能本身会波动，日常情境也千变万化，我们永远不可能做出百分之百精准的预测。

No, they do not, because as we said before, it's not possible, not only because there is some error variance always included in testing, but also because, again, executive functioning fluctuates, those daily contextual factors fluctuate, and so we cannot make a perfect prediction ever.

我也看过类似的元分析，比如莫琳·施密特或者埃奇科姆团队的研究，他们估计执行功能测试大概能解释日常表现中25%的变异。

I've seen that too from meta analyses, one that comes to mind from Maureen Schmidt or Edgecombe's group, I think suggested roughly 25% of the variance in everyday functioning is accounted for by tests of executive functions.

当然，这里面还有很多前提条件。

Of course, there's plenty of caveats.

这只是其中一个数字。

That's one number.

要衡量这些多样化的结果本来就很复杂，但总的来说，这已经占了不小的一块饼，虽然不是全部。

It's complex to measure these heterogeneous outcomes, but broadly speaking, that's a nice chunk of the pie, not all of it.

Yana，我这些年读你的论文，特别欣赏你提到的这个观点——人们总把表面效度和预测效度混在一起，全塞进‘生态效度’里，其实测试长什么样根本无关紧要。

I I wanna say, Yana, reading your papers over the years, I really appreciate this argument you make about the confusion with face validity versus predictive validity that both get smushed into ecological validity, and it it truly doesn't matter what the test looks like.

我们有些测试，看起来就像驾驶场景。

We have tests that look like, looks like a driving scene.

你翻一下题本，画面就像坐在车里、透过挡风玻璃看路，测的其实是注意力。

You know, you flip a stimulus booklet and it looks like you're behind the wheel of a car looking through the windshield and so it's a measure of attention.

你做这个测试的时候，可能会想：这测试能告诉我这个人开车怎么样。

You might do that test, you might think, this test is gonna tell me how this person will drive.

他今天或者六个月后，开车安全吗？

Are they safe to drive today or six months from now?

但这跟实际开车差得太远了。相反，如果你的病人做的是那种特别抽象的测试，比如连线测验，或者我们那些经典又可靠的神经心理测试——你打开一看，那些刺激物根本看不出跟开车有什么关系，压根儿不像从车窗往外看，但其实这根本不重要。

But that is, of course, so different from actually driving and in contrast to that, if your patient completes a test that looks very abstract like trails or one of our many tried and true neuropsych tests, you know, you open it up, you look at the stimuli, you don't see how that looks like the through the windshield of your car, but that doesn't matter.

关键在于，我们的测试得分能不能告诉你这个人今天安不安全，尤其是将来，比如中风患者康复结束后回到家，真坐上驾驶座的时候，他会不会出事故？

What matters is if the score on our test tells us anything about that person's safety today and especially in the future when they, you know, the stroke patient, when they are finished with rehab and return home and are actually behind the wheel, will they get an accident or not?

你在论文里说得特别好，真正重要的是预测效度和数据，所以我觉得——你要是说错了我纠正你——你觉得大家搞混生态效度，是因为‘表面效度’被当成万能标准了，其实不该这样。

And you make the argument beautifully in your paper is that the data, you know, the predictive criteria and validity is what matters and so I think, you know, correct me if I'm wrong, but you'd say the the confusion with ecological validity is because face validity has become the end all be all and, it shouldn't be.

是的。

Yeah.

完全正确。

Absolutely.

这种混淆还让很多人觉得，只要测试看起来像现实世界，就不需要严格验证。

And and confusion further leads to people believing that they don't need to properly validate tests.

只要测试看上去跟真实生活差不多，就不用再验证了。

So as long as the test looks like the real world, it doesn't need validation.

事实上真有人这么说过，文献里也有这种说法：‘这还用说吗？这测试当然能预测日常生活表现。’

And in fact, people have said that much, and there's literature where people say, oh, well, this is self evident, of course this test must predict daily functioning.

对，我同意。

Yep, agreed.

没错，这种逻辑跳跃完全说不通。

Yeah, that's a huge leap of logic that doesn't make sense.

我们刚才聊了聊手头那些评估执行功能的工具，其实还有不少工具是用来预测日常活动能力的。

So we've been touching on the tools at our disposal here, more for executive functions, there's also various tools in predicting activities of daily living.

咱们来具体聊聊这个。

Let's get into that a bit.

你在研究里就用了很多这类工具。

You use many of these in your research.

我们有基于表现的认知功能测试来评估日常活动能力。

We have performance based cognitive tests of ADLs.

还有自评和他评量表。

We also have self and other report forms.

当然，作为称职的临床心理学家，我们还得结合行为观察、访谈、神经诊断方法、病历回顾，和患者本人多交流。

Of course, we should be good clinical psychologists and use behavioral observations, interviews, neuro diagnostic techniques, records reviews, and we're with our patients.

就像我们之前说的，证据显示，尽管执行功能的评估方法饱受争议，但平均来看，它们在预测患者日常功能状态方面，还是比记忆、注意力或视觉空间能力的测试更有效。

Traditionally, as we've talked about, evidence shows that measures of executive functioning that are highly criticized are the best tests we have better than memory or attention or visual spatial skills on average with respect to predicting patients' functional status.

正如我们讨论过的，这些测试常被批评过于结构化、死板、不真实。

As we've talked about, these tests are criticized for being too structured, rigid, artificial.

我们都知道，现实生活中的表现更灵活，需要在无结构的环境中做决策和规划，所以我们的测试不可能完美，这很正常。

We do know real world functioning is more fluid and requires decision making, planning in an unstructured environment, hence the fact that our tests are not perfect.

稍后我们会请你说说自然情境下的评估方法，但现在，先跟我们聊聊执行功能测试和其他传统工具，它们在揭示患者日常生活能力方面表现如何？

So in a few minutes, we'll ask you to talk about naturalistic tests, but for now, tell us about tests of executive functioning and other traditional tools, in terms of their ability to reveal patients' everyday functioning.

对。

Yes.

首先，我想说，在临床实践中，我们真该重视临床访谈能带来的信息。

So, you know, first of all, I want to say that in a clinical world, I think we should pay a lot of attention to what we do in a clinical interview.

有时候我们低估了访谈的价值，总觉得非得靠某个测试出个数字，才显得更客观、更有说服力，但我觉得恰恰相反——访谈才是我们最重要的工具，因为如果患者说他能或不能做某事，我们还能继续追问、核实。

Sometimes we sort of underestimate what we can get out of an interview and we feel that we have to have a test that's gonna, again, give us a number and somehow the number's going to be more revealing than what we get from an interview, but I feel actually that that is the most important tool we have because we can follow-up with patients if they say they can or cannot do something.

我们可以请他们举出具体的生活例子，看看他们平时到底是怎么做的。

We can try to get concrete examples of their daily behavior.

我们也可以对患者的家属或其他知情者做同样的事。

We can do that same thing with significant others or other collateral sources.

所以我个人认为，要了解一个人的日常生活能力，临床访谈才是最重要的方法。

So I personally actually, above all, prefer clinical interview in terms of finding out how people function in daily life.

当然，有些情况我们需要具体展示和记录患者的日常功能，比如在诊断时需要排除某些情况——功能缺失作为诊断依据的时候。

Now there are situations where we want to demonstrate and document daily functioning, and that in my mind is more when we need it as a ruling for a diagnosis, like so lack of functioning as a ruling for diagnosis.

这时候就会用到一些行为评估工具，比如日常生活能力测试，像找零钱、写支票这些事，如果一个人连这些都做不了，那基本就能确定他确实不行了。

And that's where some of those behavioral measures, performance based measures of instrumental activities of daily living come in where, you know, if somebody can't make change or write a check or just getting obsolete, but any of those types of, you know, examples of instrumental activities of daily living, if they can't do that, then you know they can't do it.

在家他肯定也做不了。

They won't be able to do it at home.

所以这些评估手段能帮我们解释，为什么一个人会从轻度认知障碍发展成痴呆，除了认知测试结果之外，还能提供额外依据。

And so it's basically serves as a way of documenting why say somebody is converting from MCI to dementia, right, in addition to their performance on cognitive measures.

我觉得这些基于表现的评估，主要能告诉我们的是：当一个人真的做不了的时候，它特别有参考价值。

And so I feel like those performance based measures largely can tell you or are informative when somebody cannot basically do it.

但如果他能做，那我们还是不知道他回家后会不会出错——毕竟前面说过，很多情况在家是不一样的。

If they can, then you still don't know whether they're going to make mistakes at home, right, for the reasons we talked about.

还是拿驾照考试来说吧，路上开车的人，几乎百分之百都通过了考试，顶多就那么几个例外。

So get back to the example of a driving test, you know, virtually one hundred percent of people on the road, say for a few odd exceptions, have passed a driving test.

他们能开车，就是因为有驾照嘛。

That's why they're driving, that's why they have a driver's license.

但我们都清楚，路上有好司机，也有差司机，还有那种特别莽撞、对别人构成危险的，他们照样能考过驾照——因为驾照考试只看你懂不懂交规、会不会操作车辆。

But we all know there are good drivers and bad drivers out there, and there are reckless drivers who are very dangerous for the rest of us, and they pass the driving test, because the driving test tells you whether they know the rules of the road and whether they know mechanically how to operate a vehicle.

它可测不出你会不会边开边发消息，或者对突发状况反应迟钝。

It doesn't tell you whether they're going to text, they're going to inhibit responding to a test, right?

能不能判断什么时候、怎么变道才安全，这些都靠你自己。

Whether they're going to have the good judgment to plan where and how to merge into traffic.

这些事儿，一旦没人盯着，全看你的执行功能了，包括判断力、性格这些因素。

Those things, when left to their own devices, are going to depend on their executive functioning, including judgment, including personality on top of it, right?

所以从临床角度看，这些基于表现的测试，其实在这儿就失效了。

And so that's where basically those performance based measures fail us in the clinical sense.

会做一件事，不等于你能做好它。

Knowing how to do a task doesn't tell us how well you'll do it.

如果你连驾驶考试都过不了，那说明你开车确实不安全，这点大家心里都清楚。

Knowing that you can't do a test, if you fail a driving test, then you know you're not going to be safe driving.

所以测试到底在什么情况下有用，得看你想问什么问题、目的是什么、针对的是哪类人群。

So where and how those tests are useful is going to depend on the question you're asking and the purpose and the population you're working with.

说到自我报告，其实很多研究都表明，你功能受损越严重，对自己的评估就越不准确，越容易高估自己。

In terms of self report, I mean, there is plenty of literature around the fact that the more impaired you are, the less accurate you are in terms of self report, the more you overestimate.

我们还知道，执行功能越差的人，越容易高估自己的能力。

We also know that the lower your executive functioning is, the more you overestimate your abilities.

一般来说，能力越低的人，越容易高估，而能力高的人反而更靠谱——我是说，真正能力强的人。

In general, people with lower abilities tend to overestimate more than people with lower abilities and, I mean, higher abilities.

实际上有研究发现，如果你处在第八十百分位左右，你对自己的评估可能是最准的。

So actually there's research that shows that if you're about an eightieth percentile, you may be the most accurate about your abilities.

一旦超过第八十百分位，你反而可能低估自己；而能力越低的人，越容易高估。

Once you're above eightieth percentile, you may actually underestimate, whereas the lower you are, the more you overestimate.

所以，当你听病人自我报告的时候，这些因素都得一块儿考虑进去。

So you need to consider all those things when you're interpreting what a patient is telling you on self report.

说到这个，访谈的优势就在于你可以进一步追问，不只是问他们能不能做某事，而是深入了解他们是怎么做的，背后的场景是怎样的，以及他们自己是怎么描述这些事的。

In terms of, and this is again where an interview is more helpful because you can follow-up not just with asking them whether they can do something but asking them more about how they do it and getting at more of the context within which they do those things and how they're describing them.

至于旁人提供的信息，不管往哪个方向，都可能存在各种各样的偏见。

In terms of collateral sources for reporting how people do, there are so many biases in either direction that you can have.

有时候啊，家人会下意识地保护自己的亲人，替他们隐瞒问题。

Sometimes, you know, loved ones are protecting their relative.

你们有没有在住院康复科接触过有流畅性失语的病人？家属总是以为病人听得懂他们说话，其实根本不是——比如你端着饭说：‘该吃饭了，想吃吗？’病人点点头。

I don't know if you guys have ever worked like in inpatient rehab with patients with fluent aphasia where relatives invariably think the patient understands them when they do not because the patient responds to the stimuli like, here is lunch, do you want to eat?

病人说‘是’。

Patient says yes.

你看，他听得懂嘛。

Oh, see, he understands.

所以家属对病人能力的判断，往往靠不住，容易误读。

And so relatives can be very, very unreliable in terms of how they interpret what a person can or cannot do.

当然也有相反的情况，有些家属可能想把老人送走，安排到别的地方去，于是就把问题夸大，说老人什么都做不了。

Sometimes you can have the bias in the opposite direction, the relatives may be invested in getting an elderly relative placed in a, you know, some different setting, move them out of the home and may exaggerate deficits.

所以我有很多顾虑，不是说这类信息一点用都没有——比如访谈确实能提供一些帮助，但它绝不是万能的。

So I have a lot of problem, not that it can't contribute anything, again, sources like interview contributes, but it's not the be all and end all.

所以这些就是我们主要用的方法，也是我对它们效用的一些保留意见。

So those would be the main methods we use, and those would be my caveats about how useful they are.

对，你说得很有道理。

Yeah, that's good points.

没错，一个可靠又细心的第三方信息源，价值简直无可估量，但这种人真不好找。我总是会主观评估自己对这个信息源的信任度——比如他是不是天天住在家里、经常观察患者？还是说一个月才见一次？后者对我的帮助就小多了，就像你刚才说的，影响因素太多了。

Right, a reliable, observant, collateral source could be worth their weight in gold, but it's hard to get that sometimes and I'm always subjectively measuring the confidence that I have in that collateral source and whether they are in the home and observing them regularly or it could be a collateral source that sees them once a month and then they're less useful to me than if somebody sees them every day and lots of factors like you mentioned.

针对大家对我们传统认知测试（包括执行功能测试）的批评，现在学界越来越倾向于开发更贴近真实生活的评估方式，就像我们刚才聊的那样。

In response to criticisms of our conventional cognitive tests including the tests of executive functioning, There has been a push to develop naturalistic measures which resemble the real world in in some way like we were talking about.

这些方法都有很强的现实代入感，比如瑞安提到的驾驶场景测试。

They have that face validity and it ranges from driving scenes tests like Ryan had mentioned.

还有‘外出夜游’测试，以及各种各样的虚拟现实评估工具。

There's the night out test, a variety of different virtual reality measures.

这些方法大多要么模拟真实环境，比如像作业治疗师常做的厨房场景，要么用虚拟现实技术把这类场景数字化重现。

Most of these either simulate a real world environment like a kitchen in in the real world similar to what occupational therapists might do or they use virtual reality to simulate that world digitally.

关于它的进展速度。

On the pace of it.

它们本该很好地预测日常表现，因为看起来、感觉上都跟真实世界一模一样，但事实往往不是这样。

They should predict everyday functioning well because they look and they feel just like the world tests, but that isn't always the case.

很多时候都不是这样。

It's often not the case.

那你能不能跟我们说说，你在这些情境化测验上发现了什么？你觉得这些研究意味着什么？

So can you tell us what you have found in regard to these naturalistic measures and what you see as the implications of all this work?

当然可以。

Yeah, absolutely.

我特别注意到，人们搞这种情境化测验已经快四十年、快三十年了，相关论文多得数不清，各种测试都号称经过验证，可到头来，这些工具压根儿就没进过临床实践，对吧？

What I've been struck by is that people have been developing naturalistic tests for, what, forty years now, thirty years, and there are so many publications that describe different tests and quote unquote validate these tests, and yet we have yet to see these tests in clinical practice, right?

那问题出在哪儿呢？

So why is that?

于是我仔细查了查，发现大家往往就是做个测试、验证一下，然后就完了。

And so I started looking into this more carefully and it seems to me that people develop a test, validate it, and then it ends.

这很可能是因为没人重复验证过。

And it's probably because it wasn't replicated.

同一个实验室开发出一个新的测试，验证一下，发篇论文，然后就完了，再也没下文。

And so the same lab may develop a new test and validate it, right, publish a validation, and then it ends, and this doesn't go any further.

当你去翻文献，琢磨为什么会出现这种情况、为什么没人重复时，就会发现：整个领域都这样，我可没点名谁——我们都一样，对用超小样本做自然情境测试的研究睁一只眼闭一只眼，觉得能发出来就行。

And so when you start looking at the literature, like why is this happening, why aren't we replicating, You come to realize that we have been, as a field, I'm not going to point any fingers on anybody because we've all been doing it, we've been complacent in publishing studies with teensy samples that we know can't possibly be replicated for studies that use naturalistic tests.

我其实对比过自然情境测试和传统纸笔测试在验证时的科学严谨性，样本量差距大得惊人。

I actually did a comparison of the scientific rigor for validation of tests that are naturalistic versus validation of tests that are just paper and pencil tests, and there is a huge difference in the sample sizes.

对于新开发的纸笔测试，我们要求研究者必须符合严格的科学标准；但对自然情境测试，不知为啥，我们却放低了要求。

We require normal scientific rigor of researchers who want to validate a novel paper and pencil test, and we do not have those same requirements for whatever reason of people who propose to validate a naturalistic test.

所以，这种测试听起来又酷又时髦，我们潜意识里就信了——觉得它既然长得像真实世界，就一定能预测真实表现，于是对粗糙的研究方法也网开一面，这其实是个误区。

So there is, again, this sort of sexy, cool quality to it and we sort of disbelief, which is a myth, right, that the test by virtue of looking like the real world will predict real world functioning, and so we give a pass to poor scientific rigor.

这其实有一篇已经发表的系统性综述，专门分析了这些科学严谨性的指标，结果一目了然：我们整个领域确实一直这么干。

So this is a relatively published systematic review that looked at all these different indices of scientific rigor and it's very clear that that's what we've been doing as a field.

所以，也许这就是为什么我们没法重复实验，但就算研究质量不高，这些测试本身是不是至少还能预测日常生活功能呢？

So, okay, so that may be why we're not replicating, but maybe just because the research isn't great, maybe the tests still do predict daily functioning grade, right?

所以我问了自己这个问题，做了几项研究，比较了一种跟现实世界完全不沾边的传统测试，或者一个看起来跟现实毫无关系的实验任务，和一些被提出的自然情境测试。

So I asked myself that question and did a few studies where we compared a test that is nothing like the real world, so one of our traditional tests or, an experimental task that doesn't look like anything like the real world, versus some naturalistic test that's been proposed.

我们看的是那些跟自然情境测试相似的现实表现结果，然后对比这两种测试预测真实生活能力的优劣。

And looking at outcomes that look like the naturalistic test, and pitting those two tests against each other in terms of their ability to predict real world functioning.

我们选了一些很实在的指标，比如人们在家实际管理药物的情况。

We looked at pretty good outcomes, like for example, we looked at, people's actual, medication management at home.

我们每周都数他们的药片，看他们每周漏吃多少，连续跟踪了八周，这其实是个挺靠谱的指标，能清楚看出他们是不是按时吃药。

So we had, we actually counted their pills every week to see how many pills they missed in any given week, and we followed them for eight weeks, so it's a pretty good index of whether they've been taking their medications correctly or not.

我们用了一个挺巧妙的测试，叫‘药盒测试’，它跟执行功能密切相关。

And we used a clever test called the pillbox test, which correlates with executive functioning.

人们要把一堆复杂的药片放进复杂的药盒里，这个过程完全考验执行功能，跟现有的执行功能测量结果高度相关，整个过程大概十分钟就能做完。

People put complicated set of pills and complicated pillbox and absolutely taps into executive functioning, correlates with measures of executive functioning, takes about ten minutes to administer.

我们拿它跟实验室里另一个实验性测试做对比，那是我实验室用的一个电子版的‘握拳敲击测试’，同样也花十分钟左右。

We compare that to another test that is an experimental measure used in my laboratory that's basically an electronic version of Flurious fist etch pound test, also takes about ten minutes to administer.

这两个测试的可靠性差不多，其他条件都一样，唯一的区别就是一个是自然情境的，一个不是。

Both tests had similar reliability, so everything being the same other than one was naturalistic and one was not.

结果发现，我们那个握拳测试对服药管理的预测效果比药盒测试还要好，说明它在药盒测试之外还有额外的预测价值。

And turns out that our fistage palm version predicted medication management better than the pillbox test did, basically showing incremental validity above and beyond the pillbox test.

那药盒测试本身预测得怎么样？

So did the pillbox test predict?

是的，它确实能预测。

Yes, it did.

但没那么准。

But it wasn't better.

它反而更差。

It was poorer.

不只是没跟它打平，对吧？

Not only was it, it wasn't even equal, right?

它比另一个测试表现得还差。

It was poorer than the other measure.

所以当你验证这些自然情境测试时，发现它们跟执行功能指标或某些结果指标有关联，这还远远不够，对吧？

And so when you validate some of these naturalistic tests and show that they correlate with measures of executive functioning or they correlate with some outcome measure, that's not sufficient validation, right?